Abstract
For the first time it was demonstrated that an osteoinductive calcium phosphate-based putty is effective in the restoration of complex maxillofacial defects. In these defects, adequate mechanical confinement by multiple bony walls and osteoconduction from multiple surfaces are usually lacking. This study compares the efficacy of a microstructured beta-tricalcium phosphate (β-TCP) putty with autologous bone for the repair of alveolar cleft defects. A total of 10 Dutch milk goats were operated on in a split-mouth study design in which two-wall bony alveolar clefts were created and successively repaired with autologous bone (the gold standard) at one side and β-TCP putty at the other. After 24 weeks of implantation, histomorphometric and micro–computer tomography analyses proved that the β-TCP putty group showed equal bone quality and volume to clefts reconstructed with autologous bone. In addition, surgical handling of the putty is superior to the use of calcium phosphates in a granular form. Therefore, the results of this study open a clear trajectory for the clinical use of β-TCP putty in the reconstruction of the alveolar cleft and other challenging two-wall bony defects.
In patients with cleft lip and palate, alveolar bone grafting is paramount. It provides continuity of the dental arch and closure of the oronasal communication, facilitates tooth eruption of teeth adjacent to the cleft and successive orthodontic tooth movement, restores the support of the nasal base, and provides stability of the maxillary arch (Boyne and Sands, 1972; Bergland et al., 1986). Until now, for the repair of alveolar cleft defects, autologous bone grafts are the gold standard (Enemark et al., 2001). Because these bone grafts are harvested from either the iliac crest or the mandibular symphysis, donor site morbidity is introduced (Swan and Goodacre, 2006; Weibull et al., 2009). To bypass this cause for postoperative pain and to reduce operating time, the use of synthetic bone substitutes is a logical next step.
To be suitable for alveolar cleft repair, the selected bone substitute should be resorbable, meaning that it should be replaced by vital bone in time (creeping substitution). In this way, tooth eruption, orthodontic tooth movement through the grafted area, and root stability of the erupted tooth are guaranteed. These requirements have strict implications on the degradation kinetics of the bone substitute selected. For example, if resorption is too slow, orthodontic tooth movement will be blocked. On the other hand, if resorption is too fast, insufficient volume of scaffold will be provided to the growing osteoblasts, as such frustrating bone formation, leaving an unfilled alveolar cleft.
As an alternative for autologous bone grafts, osteoinductive calcium phosphate scaffolds (OCPSs) seem to be a promising candidate for alveolar cleft repair. OCPSs are biomaterials able to induce bone formation when implanted at heterotopic sites, an ability known as osteoinduction. This phenomenon has been recognized for 30 years (Barradas et al., 2011), and the superiority of OCPSs has been demonstrated above osteoconductive calcium phosphate ceramics in clinically relevant models (Habibovic et al., 2006; Yuan et al., 2010). Although the osteoinduction mechanism is not yet completely unraveled, the relationship between the physical and chemical features of the OCPS material and the osteogenic differentiation of human mesenchymal stem cells has been demonstrated in vitro (Yuan et al., 2010; Davison et al., 2014).
Furthermore, osteoinductive calcium phosphates do not have the drawbacks of the existing osteoinductive alternatives, that is, autograft (i.e., two surgical procedures, donor site morbidity, prolonged operating time) and bone morphogenetic proteins (i.e., systemic effects leading to severe side effects, higher risks in the pediatric population [Epstein, 2013], expensive [Dickinson et al., 2008]).
In a recent pilot study alveolar defects in six patients with cleft lip and palate were successfully reconstructed with β-TCP granules (de Ruiter et al., 2015). However, the dimensional stability of the reconstructed sites was found to be suboptimal because of the free motion of the granules. The alveolar cleft defect typically is a two-wall bony defect in which mucoperiosteal flaps are sutured in two layers to create both a new nasal floor and a continuous oral mucosa. It is very challenging to create a watertight closure of these mucoperiosteal flaps to retain the graft within the allocated area and to prevent microorganisms from infecting the graft. In the alveolar cleft defect, it is of paramount importance that the applied bone graft or bone substitute remains in place.
The addition of these β-TCP granules with a binder into a putty is a significant improvement for both the surgical handling and the initial dimensional stability of the reconstructed area, provided that the binder does not hinder bone formation (Bohner et al., 2010; D'Este et al., 2013). Davison et al. (2012) showed that a blend of carboxymethyl cellulose in glycerol (CMCG) is a suitable binder formulation in combination with osteoinductive β-TCP granules. The CMCG binder is inert, biocompatible, and ensures the initial containment of the β-TCP granules in the defect. With a rapid clearance kinetics (<4 hours in vitro, <48 hours in vivo), the CMCG binder allows the rapid exposure of the β-TCP microstructure surface for optimal bone growth. Furthermore, the water-free nature of CMCG prevents the alteration of the microstructured surface upon storage. Therefore, the resulting putty can be readily applied at the surgical site.
In the present animal study, microstructured β-TCP granules, embedded in a CMCG matrix, were used as an alternative grafting material for alveolar cleft repair in a goat bilateral alveolar cleft model. It was hypothesized that the putty would perform at least equally to the autograft gold standard concerning residual bone volume and quality and that surgical handling would prove superior to the use of regular β-TCP granules.
Materials and Methods
Study Design
A goat model was used that has been previously described by the authors (de Ruiter et al., 2010). The study and its protocol were approved by the Dutch Animal Care and Use Committee (DEC-UMC 2013.III.02.013). A split-mouth study was conducted on 10 female adult Dutch milk goats (capra hircus; Heythuysen, the Netherlands) in which bilaterally created alveolar clefts were repaired using β-TCP-CMCG on one side and autologous bone harvested from the iliac crest on the other side. Appointment of the β-TCP-CMCG grafting side was randomized. All goats were aged between 36 and 38 months, ensuring that no deciduous teeth remained and to prevent sample bias as a result of age.
Surgical Procedure
Surgery was performed under general intravenous and inhalation anesthesia. For extraoral disinfection, a 1% iodine in 70% ethanol solution was used. Intraoral disinfection was achieved with 0.12% chlorohexidine. After extraction of the left and right maxillary second premolars, the buccal and palatal mucoperiosteal flaps were raised. Consecutively, two-wall bony defects of approximately 1 cm3 were created in the maxilla by removing all buccal, palatal, and nasal bone. The nasal mucosal layer was left intact. The bone defect on one side was filled by injecting the β-TCP-CMCG putty without a need for additional handling, for example, reconstitution, hydration with water, or component mixing. The defect on the other side was repaired with autologous cancellous iliac crest bone chips. The mucoperiosteal flaps were closed in a tension-free manner using resorbable sutures (Vicryl 3-0, Ethicon, Brussels, Belgium; Figs. 1 and 2).
Schematic drawing of the surgical procedure. Region of interest in the goat maxilla is indicated with a dotted line. The defect site is located between the first (P1) and third (P3) premolar and is cranially confined by the nasal floor (N). The defect was successively repaired with grafting material (G).
Intraoperative view. Surgical creation of the alveolar cleft defect between the first and third premolar after extraction of the second premolar and removal of bone up to the nasal mucosa (A). One defect site was filled with β-TCP-CMCG putty (B) and the contralateral with autologous iliac crest bone chips (C). Mucoperiosteal flaps were raised and the defect sites were closed in a tension-free manner (D).
Bone Substitute Preparation
The putty was a kind gift from Xpand Biotechnology BV (Bilthoven, The Netherlands). The microstructured β-TCP (composition: 98% β-TCP, 2% Hydroxyapatite) ceramic particles were made by wet precipitation of apatite powder (calcium phosphate [Ca/P] ratio 1.5), followed by green body H2O2 foaming and sintering at 1050°C, as previously described (Yuan et al., 2001). The desired TCP granule fraction of 150 to 500 μm was collected after sieving (Retsch, Haan, Germany) and cleaned with ethanol, acetone, and deionized water. Total porosity was found to be 70% by mercury intrusion testing (Micromeritics, Aachen, Germany).
The CMCG binder was prepared by blending 4% in weight carboxymethyl cellulose (CMC; Cekol, CPKelco, Atlanta, GA) with glycerol (Sigma-Aldrich, St. Louis, MO). CMC is a cellulose derivative that is physiologically inert, hypoallergenic, nontoxic, nonmutagenic, and noncarcinogenic. CMC has a flaky structure. It is hydrophilic and acts as a thickener and suspending agent. CMC is mostly combined with water for form a binder, but it can form also a thick gel in combination with anhydrous glycerol, allowing long-term preservation of the osteoinductive β-TCP microstructure.
The micro structured β-TCP granules were gently blended at 80°C with the CMC and glycerol matrix using a spatula and packed individually. The resulting putty contained 50% granules and 50% CMCG binder. The binder occupied the intergranular space, and thus in volume, the granules occupied >90% of the putty volume so that the bone defect is essentially filled with the microstructured β-TCP after CMCG clearance. The putty was terminally sterilized by E-beam at 25kGy (Synergyhealth, Radeberg, Germany).
Animal Care
Starting 1 week preoperatively to 3 weeks postoperatively, all goats were fed with premoistened ground chunks of beet pulp mixed with lucerne pulp (De Heus Diervoeders, Rijsbosch, Beusichem, The Netherlands). This was precured to prevent wound dehiscence. During the remainder of the pre- and postoperative periods, all goats were fed with hay and regular ground chunks (Arie Blok Diervoeding, Woerden, The Netherlands). At 6 months postoperatively, all goats were sacrificed by means of an overdose of pentobarbital (Euthesaat, Organon, Oss, The Netherlands).
Radiographic Assessment
The specimens retrieved from the sacrificed animals were scanned using a micro–computer tomography (CT) scanner (Quantum FX Micro-CT; PerkinElmer, Waltham, MA). Samples were scanned at a 120 μm voxel size (90 kV, 180 μA current, 1.0 mm aluminum (Al)/0.25 mm copper (Cu) filter, and 0.5° rotation step, 20-minute scan). All scans were blinded, and residual bone volumes were calculated using a DICOM viewer (Osirix; Pixmeo, Geneva, Switzerland). The region of interest consisted of the area between between the first and third premolar in the anteroposterior plane and mediolateral plane. The defect was measured up to the nasal floor in the craniocaudal plane. Residual bone volume was quantified after the segmentation of calcified tissue from noncalcified tissue using a global threshold. This global threshold was determined based on visual inspection and was kept constant for all scans. The segmentation of residual bone volume was performed manually using a free-form tool. After the segmentation of each slide within the region of interest, the total residual bone volume was calculated.
Histological Procedure and Assessment
The specimens were first fixed for 1 week in 4% formalin solution (VWR-Prolabo, Radnor, PA); thereafter dehydrated using the Milestone TT Mega histoprocessing microwave in 80% ethanol (VWR-Prolabo), JFC solution (Milestone; Klinipath, Duiven, The Netherlands), and two steps of 100% ethanol. They were embedded in K-plast polymethyl-methacrylate (MMA) (L.T.I., Bilthoven, The Netherlands), sectioned with a slice thickness of 50 (im using a Leica SP1600 saw (Leica Microsystems, Rijswijk, The Netherlands) with a diamond saw blade (Saint-Gobain Diamantwerkzeuge, Norderstedt, Germany), and stained with methylene blue and basic fuchsin (Sigma-Aldrich). Histomorphometric analysis was conducted as widely and previously described (Habibovic et al., 2006; de Ruiter et al., 2010; Yuan et al., 2010; Barbieri et al., 2011; Davison et al., 2012, 2014). Sections were digitalized using the Konica Minolta Dimage II slide scanner (Marunouchi, Chiyoda, Tokyo), and histomorphometry was performed to determine the amount of bone in the available area. Using Photoshop Elements software (Adobe Systems, San Jose, CA), a region of interest was selected between the two premolars, the nasal and palatal mucosa, representing the defect. In the region of interest, the total amount of pixels was determined as well as the amount of bone and residual material (if applicable) by pseudocoloring.
Statistical Analysis
A paired-samples t test was conducted to compare the differences in residual bone volumes between the β-TCP-CMCG and iliac crest graft as assessed from the micro-CT scans (SPSS version 21; IBM, Chicago, IL). A second paired-samples t test was conducted to compare the percentages of bone formation on the histological sections. A P value <.05 was considered as statistically significant. All data are represented as means (standard deviations).
Results
Clinical Observations
Surgical handling of the putty showed to be very effective. Although the surgically introduced maxillary defects were difficult to reach because the goats were in a prone position, the putty was easily administered with a syringe and remained in place. All intraoral operation sites healed successfully without any signs of wound dehiscence or infection. After sacrificing all goats, one of the animals (goat number 6, β-TCP-CMCG side) appeared to suffer from a mobile first premolar showing signs of local periodontitis because of an impacted hay splinter, but it was not excluded from the study.
Micro-CT Scans
Volumes of the restored bone defects on both sides of the maxilla are depicted in Figure 3. A paired-samples t test did not show any statistically significant difference between cleft sites grafted with β-TCP-CMCG or autologous bone graft (P = .34). All samples showed continuity of the maxillary bone at the defect site and formation of normal cortical and cancellous bone. The mean residual volume was 0.71 (0.2) cm3 for the β-TCP-CMCG group and 0.64 (0.27) cm3 for the iliac crest group.
Residual volumes of the reconstructed defects are shown for all animals at 6 months postoperatively. The beta-tricalcium phosphate carboxymethyl cellulose in glycerol (β-TCP-CMCG) group and the iliac crest control group are depicted per animal.
Histological Sections
At histological evaluation after 6 months, the tissue responses were similar for all implanted graft materials: all tissues surrounding the implants were normal, and regular inflammatory reaction was observed on the β-TCP-CMCG group comparable to β-TCP as reported by Davison et al. (2012). Microscopically, no cytotoxic effects were detected in the host tissue surface in contact with the implanted material. There were no significant differences in terms of biological response of tissues surrounding the implants between the β-TCP-CMCG group and autograft groups, indicating the biocompatible nature of the CMCG binder.
All samples showed bone regeneration yielding bone tissue with an architecture resembling that of natural bone (Fig. 4). Histomorphometric analysis of the area percentages of newly formed trabecular and cortical bone structures revealed an average of 28.9% (5.5) in the β-TCP-CMCG group and an average of 28.6% (7.9) in the autologous bone group (Fig. 5). There was no significant difference in the bone area percentages between the two groups (P = .91). On average, only 0.1% of biomaterial was discernible in the β-TCP-CMCG group, with a maximum percentage of 0.6% in goat 1.
Bone formation in an alveolar cleft grafted with beta-tricalcium phosphate carboxymethyl cellulose in glycerol (left) and in the contralateral alveolar cleft defect repaired with an autologous iliac bone graft (right). Both sections show a trabecular bone architecture resembling a natural maxillary bone structure. Dotted line represents the region of interest. P1, first premolar; P3, third premolar; N, nasal floor; G, gingiva.
Area percentage of newly formed bone in the defect site from histomorphometric analysis. The beta-tricalcium phosphate carboxymethyl cellulose in glycerol (β-TCP-CMCG) group and the iliac crest control group are depicted per animal.
Discussion
Newly formed bone with an architecture resembling natural maxillary bone was found in all reconstructed defects repaired with both autologous iliac crest bone and β-TCP-CMCG. The two grafting materials showed comparable results. The surgical handling of the putty was excellent and remained in place adequately.
Animal Model Considerations
In a previous study, it was shown that the goat model is suitable to study alveolar cleft repair (de Ruiter et al., 2010). The created defect is similar in size as in alveolar cleft patients, and the goat metabolism is more equivalent to the human situation than phylogenetically less kindred species such as rodents in which bone metabolism is higher and defect sizes are much smaller (van der Donk et al., 2001; Muschler et al., 2010).
It has been extensively shown in the human analogy of the alveolar cleft defect that the bony defect is not always of critical size (Hellquist et al., 1983). In more than 50% of human cleft patients, a very small bony bridge is established when the alveolar cleft is closed using only the surrounding mucosal tissue without implanting a bone graft. Therefore, in this study, a sham group was not studied. Thus, because we are not trying to repair a critical size bone defect, we have not considered it worthwhile to create a sham group, taking into account the reduction in use of laboratory animals. Accordingly, Koole et al. (1991) showed that in a sheep cleft defect of similar size and locus, the sham group (in which no bone graft was used) showed regeneration of the defect. However, this occurred in a severely delayed way, with insufficient bone volume when compared with the grafted subjects. As a critical note to the study design, carry-over effects from one operated site to the contralateral one and vice versa may induce bias in split-mouth studies (Hujoel and DeRouen, 1992).
Surgical Considerations
Surgery consisted of a one-stage operation in which the defects were bilaterally created and immediately restored. Furthermore, in the human situation, a complete alveolar cleft is also compromised by the presence of an oronasal fistula, an unwanted communication between the oral and nasal cavity. As a result of the one-stage character of the surgery, while creating the alveolar cleft defect, the nasal mucosa was left intact because piercing and suturing it in the same session through the created alveolar defect would be surgically far more unreliable than in a human. A remark on the translatability of the surgical model can be made on these two issues because the closure of a chronic oronasal fistula in a human is a more challenging matter than the healing of an alveolar wound that did not suffer from a nasal communication, as was created in the animal model. However, because of the animal ethical standards in the Netherlands, a model in which two consecutive surgeries are performed for this procedure is not allowed.
The putty-like structure of the β-TCP-CMCG provided optimal surgical handling. CM CG was chosen as a carrier for the β-TCP because Davison et al. (2012) described 100% granule retention and orthotopic bone bridging, as well as ample ectopic bone formation, in an in vivo dog's femur model. However, contrary to the bony defect that was investigated in our study, the defect was only monocortical, consisted of five bony walls, and was only 5 millimeters in diameter. In Davison's study, β-TCP-CMCG showed a slightly delayed onset of bone formation when compared with β-TCP granules only. However, after 12 weeks, the amounts of bone formation were equal. CM CG is a rapidly dissolvable carrier with a dissolution time of <4 hours in vitro and <48 hours in vivo. Barbieri et al. (2011) showed that carriers dissolving rapidly (i.e., within 48 hours) do not inhibit the osteoinductive potential of the β-TCP granules. The choice for a nonaqueous carrier has been made because of the prolonged shelf-life when compared with aqueous carriers (LeGeros, 1993).
Histological and Radiographic Considerations
Volumetric and histomorphometric analyses showed no significant differences between β-TCP-CMCG and autologous bone harvested from the iliac crest. All samples showed bone regeneration resulting in a normal bony architecture. In the β-TCP-CMCG samples, scarcely any β-TCP remnants were visible. When sparse remnants of the β-TCP were studied, osteoinductive properties of the scaffold could be seen (Fig. 6). These findings concur with our earlier published data on osteoinductive properties of similar microstructured TCPs (Yuan et al., 2010). In most samples, there was a considerable amount of decrease in volume of the restored defect 6 months postoperatively, regardless of the grafting substance. This is also observed in humans when the alveolar cleft bone is not functionally loaded by a tooth that has erupted or orthodontically maneuvered into the newly formed bone (de Ruiter et al., 2015). Therefore, this volume decrease was attributed to a lack of functional loading.
Osteoinduction by beta-tricalcium phosphate carboxymethyl cellulose in glycerol putty in the maxillary cleft of goats at 6 months. At low magnification (left), induced bone tissue (red) surrounds partially degraded TCP particles (brown) in the soft tissue (light blue) of the goat maxilla (scale bar left = 250 μm). A: At higher magnification, a multinucleated osteoclast-like cell (white arrowhead) resorbs the material (white stars) adjacent to newly formed bone (black star). B: Elsewhere, cuboidal osteoblasts (black arrow heads) lay down new bone (pink) adjacent to an osteocyte (white arrow) in its lacuna. Scale bars A, B = 25 μm.
Conclusion
β-TCP-CMCG putty allows for excellent surgical handling in the repair of alveolar cleft defects in a goat model. When compared with the alveolar cleft grafting of a similar defect using granules only, the handling and initial form stability of β-TCP-CMCG putty are superior (de Ruiter et al., 2010). After 6 months of healing, the grafted sites with either β-TCP-CMCG putty or autologous bone show a continuous maxillary arch with normal bone architecture. Hardly any remnants of β-TCP are discernible in the histological specimens. The results of this study open a clear trajectory for clinical use regarding the repair of the human alveolar cleft with β-TCP-CMCG putty.
Footnotes
Acknowledgments.
Special thanks to Linda van Leeuwen (Xpand Biotechnology BV) for her extensive support during surgery and histology.