Biocompatibility of fiber-reinforced composite (FRC) and woven-coated FRC: an in vivo study

To investigate biocompatibility and bone contact area of FRC and woven-coated FRC (FRC-C) in rats. Sixty rats were allocated to three groups: FRC (n=20), FRC-C (n=20), and control group (n=20). Subgroups were determined as 4th (n=10) and 12th weeks (n=10). The specimens were placed in the femur of rats. In the control group, the bone defects were left empty and sutured. Four and 12 weeks after implantation, the rats were sacrificed. Histopathological examinations were performed in a semi-quantitative manner. Twenty rats (n=20) were used for scanning electron microscopy (SEM) examination. Bone contact surfaces were calculated in SEM analysis. A chi-square test was performed to analyze the data. No statistical difference was detected between the 4th and 12th weeks in the quality of bone union. Quality of bone union was lower in FRC compared to the control group in the 4th week (p=0.012) and the 12th week (p=0.017). The periosteal reaction at the 12th week was lower in FRC than in the control group (p=0.021). Bone contact of FRC and FRC-C was 85.5% and 86.3%, respectively. FRC and FRC-C were biocompatible and showed no inflammation. The woven coating did not increase the quality of bone union and bone contact area, while not reducing biocompatibility. The biocompatibility and good bone response of the woven glass fiber net were demonstrated to have the potential as a scaffold for the augmentation of alveolar bone deficiencies and the reconstruction of maxillofacial defects.


Introduction
The survival of oral implants is related to the formation and maintenance of a proper soft or hard tissue response and osseointegration [1]. The superior properties and osseointegration process of titanium or titanium alloy (Ti6Al4V) make it widely used as a dental implant; however, there are several drawbacks of these materials, and one of them is its elastic modulus [1,2]. Metallic implants (110 GPa) have a higher elastic modulus, approximately ten times higher, compared to bone tissue (15 GPa) [2]. The mismatch between the elastic modulus of the vital tissue and implant material causes improper stress distribution that results in resorption and inadequate osseointegration [3]. As a consequence, this led to the development and use of fiber-reinforced composites (FRC) in dentistry that has a more similar elastic modulus to the bone [4,5]. The soft tissue response of metallic implants has great clinical importance considering the biological, mechanical, or esthetic consequences of the treatments. It has been shown that Ti +4 ion can cause soft tissue atrophy and thus the exposure of the implant material [6]. In addition, because of the metallosis or immunological effects related to metallic nanoparticles released from the implant, the development of non-metallic implants has been accelerated [7].
The reinforcement of resins with either short or long fibers has been described in the literature for more than 40 years [4]. With the introduction of new technologies, for FRC devices there are now new open fields of research both in laboratory and clinically [5]. The microstructure of FRC consists of collagen matrix as an organic component and hydroxyapatite crystals as an inorganic component, which is still considered one of the most biocompatible materials for bone tissue [8]. Its adequate mechanical or biological properties, high impact resistance, and magnetic-resonance compatibility are promoted its use as a dental implant or reconstruction material [9,10]. Moreover, FRC allows forming of a monoblock-like structure including the bone and implant due to its elastic modulus of 15-20 GPa [1,11]. On the other hand, biological properties of the FRC were widely studied by cell culture or in vivo research and demonstrated proper cell response, similar osseointegration to titanium [1,8], and good biocompatibility without inflammation or toxicity [1,2,12,13].
FRC has recently been used in various medical purposes such as orthopedic and cranioplasty surgeries, or maxillofacial reconstructions [14]. To reconstruct the maxillofacial bone defects associated with infection, malignancies, or trauma and that require mechanical and biological adequateness is a challenge. It has been reported that the in vitro mechanical durability of FRC in the reconstruction of the maxillofacial defect was sufficient [15]. According to the literature, titanium is still the best option for dental implants due to its optimum osseointegration properties; on the other hand, when used as a reconstruction material in the augmentation of irregular maxillofacial bone defects, it is rigid enough to damage the surrounding vital tissues in the case of excessive force or trauma [3,16]. Despite the favorable elastic modulus and mechanical properties of polymer-based FRC, its potential use as a mesh material for bone augmentation in the case of inadequate bone volume observed as a consequence of the edentation and hampered by the proper placement of the implant can be considered.
Osseointegration is a direct connection between the bone and the implant material that acquire structural and functional harmony [1]. The surface modification of the implant material is an option to enhance the osseointegration process [4,17]. The surface of porous materials enables the growth of blood vessels and mineralization, the so-called osteoinduction mechanism. The roughened surface of dental implants increases the quality of osseointegration and the shear bond strength between the material and the bone by allowing vital tissue ingrowth [17]. Not only from the mechanical perspective, but also it has been shown that the optimal roughness (approximately in Ra value of 1-2 μm) promoted osteoblast differentiation [18]. In the literature, the roughness of the materials was created by coating or the modification of the surface such as grit blasting, acid etching, calcium phosphate coating, or bioactive glass (BAG) embedding [1]. A recent previous study used titanium as a body and coating material in fiber form and reported increased bone formation [19]. In the present study, we used an FRC body, and a coating material consisting of woven polymer impregnated E-glass fiber and investigated the bone reaction. Although there are several in vitro and in vivo data based on the fiber-reinforced composite material, there is limited information about the mesh integrated FRC. The present study aims to investigate the osseointegration and biocompatibility of FRC and woven-coated FRC. Our first null hypothesis was that there would be no difference between the biocompatibilities of FRC and woven-coated FRC, and the second null hypothesis was that there would be a higher bone contact area in woven-coated FRC compared to FRC.

Study design and groups
The study protocol was approved by the Animal Ethics Committee of Cumhuriyet University (protocol number; 124/05.06.2008). The protocol of our study was accomplished in accordance with the ARRIVE guidelines (Animal Research: Reporting of In Vivo Experiments). This study was performed on 80 healthy and skeletally mature male Wistar albino rats (Rattus norvegicus albinus), 4 and 6 months old, with an average weight of 210 g. The animals were obtained from the same facility, the vivarium of the faculty of veterinary, in order to decrease genetic variability. All animals were housed in proper individual cages with climate-controlled optimum conditions (temperature 21±2°C, 12-h light and dark cycle, 40-60% humidity) and proteincontained granular food and water provided ad libitum.
Power analysis of pilot data was calculated. The primary endpoint hypothesis was that hard tissue inflammation in study groups (FRC and CFR-C) was non-superior to the control group. The calculations were carried out using the software of G*Power 3.1 (Heinrich-Heine-Universität, Düsseldorf, Germany) with an effect size of 0.20, at the significance level of α=0.05, and a power of 0.80. Each group required 10 samples to obtain the signified values. A total of 80 rats were used for our study. For the histopathological examination, the rats (n=60) were allocated to three main groups according to the material: the FRC group consisted of only FRCs (n=20), the FRC-C group consisted of FRC with woven fiber coat (n=20), and the control group (n=20). Every main group was divided into two subgroups (n=10) according to the 4th and 12th weeks of the sacrification period. Twenty rats were used for scanning electron microscopy examination for each main group: FRC (n=10) and FRC-C (n=10), and the sacrification period was adjusted to the 12th week.

Material preparation
The properties of materials used in our study are listed in Table 1. The light polymerized FRC materials were prepared 1.5 in diameter and 3 mm in length. The specimens were measured three times to confirm and standardize the dimensions using a digital caliper (Vernier, Insize Co, Sao Paulo, Brazil). The polymerization of specimens on both sides was carried out for 40 s using a halogen hand curing unit (Optilux 501, Kerr, Danbury, CT, USA) according to the manufacturer's instructors (the curing parameters: the light intensity of 930 mW/cm 2 and the wavelength of 400-505 nm) [12]. FRC-C specimens were prepared by coating a woven polymer-monomer gel impregnated E-glass fiber laminate (EverStick Net, StickTech, Turku, Finland) in thickness of 0.06 mm onto the FRC body. Bidirectional woven fiber consists of silanated E-glass fiber and bisphenol-A diglycidylether methacrylate (Bis-GMA) and polymethylmethacrylate (PMMA) as an organic polymer matrix (composition: SiO 2 55%, CaO 23%, N 2 O 3 15%, and B 2 O 3 6%). Before the surgery, all specimens were sterilized for 20 min with an autoclave (Euroklav 23 VS, Melag Medizintechnik GmbH, Berlin, Germany) at 121°C and 0.1 MPa [20].

Surgical procedures
All surgical procedures were performed under sterile conditions. Any antibiotic medication was not given to any rats before surgery. The rats were anesthetized with an intramuscular injection of ketamine (Ketamidor 10%, Richter Pharma AG, Wels Austria) at a dose of 1 mL/kg. The rats were placed and fixed in the supine position. The anterolateral aspect of the right leg of the rats was shaved, and antisepsis was attained with 5% povidone-iodine solution (Batticon 10%, Adeka, Istanbul, Turkey). On the flexion position of the leg, linear skin incisions of approximately 2.5 mm were performed and periosteal flaps were elevated to expose the femur using a periosteal elevator. A cylindricalshaped defect (Ø 1.6 mm) was performed at a height of 3 mm to the anterior surface of the proximal diaphysis of the femur using a stainless-steel dental drill under excessive cooling with physiological saline solution. After the drilling procedures were completed, the bone defects were irrigated with physiological saline solution to remove bone fragments. The FRC and FRC-C specimens were inserted in the bone defects of the rats belonging to the study groups. For the control group, the bone defects were ensured to fill with blood clots and were closed without any intervention. No fractures on the femurs were observed on the pre-and postoperative examination. After the insertion of the specimens, the periosteal flaps were sutured with 4.0 chromic catgut (Ethicon, Somerville, NJ, USA), and the skin was sutured with 3.0 silk (Ethicon, Somerville, NJ, USA). All animals were kept under surveillance to be sure of the recovery of anesthesia effects before returning to their cages. For the prophylactic purpose, benzylpenicillin/dihydrostreptomycin was administered subcutaneously (Tardomycel, BayerVital, Germany) in a dose of 2.5 mL every 48h for 7 days. Postoperative analgesia was provided with carprofen (Rimadyl, Pfizer, USA) in a dose of 4 mg/kg via subcutaneous injection every 8 h for 7 days. The sutures were removed on the 8th day, and the surgical areas were cleaned with a 5% povidoneiodine solution. At the end of the 4th and 12th weeks after implantation of the specimens, the rats were sacrificed by a lethal dose (200 mg/kg) of pentothal sodium via intraperitoneal injection. After the sacrification procedures, the femur block samples containing the implants and surrounding tissues were dissected and prepared for histopathological and scanning electron microscopy examination.

Histological evaluations
The harvested femur blocks were stored and fixed in a 10% formaldehyde solution (International Organization for Standardization 10,993-6:2007) at a pH of 7.0 for 24 h. All samples were decalcified with 10% formic acid. After the decalcification process, the samples were divided into two pieces to expose the contact line of bone and the implant materials. The specimens of bone fragments were embedded with paraffin. The sections of 0.5 μm in thickness were prepared parallel to the long axis of the implant materials and stained with hematoxylin-eosin. The samples were deparaffinized before the histological analysis.
Histological examination was performed on the 10 sections of the 10 different slides of every study group using light microscopy (Zeiss Axioskop, Zeiss, Oberkochen, Germany) at 40×, 100×, and 200× magnification. The numbering of samples was performed randomly, and the histological analysis was performed by one pathologist blinded to the data of study groups. The histological analysis of samples was performed by evaluating the parameters such as the quality of the union, osteoblastic activity, osteoclastic activity, and soft tissue inflammation. The data obtained from histological analysis were calculated in a semi-quantitative manner by the 0, 1, and 2 scores for inflammation, osteoblastic, osteoclastic, and periosteal reaction (0: no reaction, 1: light or moderate reaction, 2: high reaction), and the 0-4 scores for the quality of bone union (0: no union, 1: fibrous union, 2: fibrocartilaginous or cartilage union, 3: mineralizing cartilage and bone union, 4: mature bone union).

Scanning electron microscopy
A detailed investigation of the surface characteristics of the samples was performed using scanning electron microscopy (SEM) at × 52, × 75, × 200, and × 460 magnifications. The samples reserved for scanning electron microscopy examination were sectioned using a diamond saw (Minitom, Struers, Denmark). The samples (5×6 mm) were stored in 2.5% glutaraldehyde for 2 h, and subsequently, washed with Sorensen's phosphate buffer. For the secondary fixation process, 1% osmium tetraoxide was used for 2 h, and Sorensen's phosphate buffer was used again for the irrigation of the samples. The samples were dehydrated with ethanol ranging from 45 to 95% and dried using the critical point drying method. SEM analysis was performed. To obtain the quantitative values from the SEM images, the measurements were performed using Corel Draw 13 software (Corel, Ottawa, Canada). The millimetric value of the distance between the outer lateral border of bone and bone marrow (a) and the millimetric value of the area of the FRC implant that contacts the bone (b) were calculated. The percentage of FRC contact surfaces was calculated as the ratio of the (b/a)×100.

Statistical analysis
Statistical analysis was performed using SPSS software (IBM SPSS Inc., Chicago, USA). To statistically analyze the data of SEM and histopathology belonging to FRC and FRC-C groups, chi-square (Fisher's exact test) was used. The level of statistical significance was set at p<0.05.

Results
No postoperative complications were observed in animals. The implantation process was well tolerated by all rats. There were no clinical symptom and no eventful situation of the general status. Skin healing was good. During the macroscopic examination of the animals, no necrosis, granulation tissue, hemorrhage, or other signs of rejection were detected around the specimens. The scores of the quality of union, osteoblastic activity, osteoclastic activity, soft and hard tissue inflammation, periosteal reaction, fibrosis, and the surface of bone contact belonging to FRC, FRC-C, and control groups are demonstrated in Figs. 1 and 2, and Tables 2 and 3. In histological examination, the implant materials are surrounded by a fibrous capsule. Collagen fibers, vascular elements, and fibroblasts were followed. In the bone tissue, osteocytes in the matrix and osteoblasts were observed. No bone necrosis was observed in any of the samples at the 4th and 12th weeks. New forming bone areas were observed in histological examinations of the FRC group in the 4th week (Fig. 3). In FRC, FRC-C, and control groups, there was no statistical difference in the quality of bone union between the 4th and 12th weeks (p>0.05). Regarding the quality of bone union, at the 4th and 12th weeks, there was no statistical difference between FRC and FRC-C, likewise FRC-C and the control group (p>0.05). However, there was a statistical difference between the FRC and control group; the quality of bone union was higher in the control group (p=0.012 for the 4th week, p=0.017 for the 12th week).
No statistical difference was detected in the osteoblastic reaction, osteoclastic reaction, hard or soft tissue inflammation, fibrosis, and periosteal reaction between the 4th and 12th weeks in FRC, FRC-C, and control groups (p>0.05). According to statistical analysis, at the 4th and 12th weeks, there was no difference between FRC, FRC-C, and the control groups in the osteoblastic, osteoclastic reaction, hard or soft tissue inflammation, and fibrosis (p>0.05).
In the periosteal ration, at the 4th week, there was no statistical difference between FRC, FRC-C, and the control groups (p>0.05). However, in the 12th week, while no statistical difference was observed between FRC and FRC-C, similarly in FRC-C and the control group (p>0.05), there was a statistical difference between FRC and the control group (p=0.021). The periosteal reaction was statistically higher in the control group compared to the FRC group (p=0.021).
According to the SEM analysis, new bone formation was detected at the surface of the samples of FRC and FRC-C groups (Fig. 4). There was no bone necrosis and resorption. In quantitative SEM data, no statistical difference was observed between the FRC and FRC-C groups (p>0.05). The surface of bone contact in FRC and FRC-C was 85.5 (±9.8) % and 86.3 (±11.7) %, respectively (Table 3).

Discussion
In the present study, we examined the osseointegration and soft and hard tissue response of FRC and wovencoated FRC at the 4th and 12th weeks of implantation and reported new bone tissue formation, no inflammation, and high bone contact area. From a classic perspective, inert and tolerable material is considered biocompatible; however, the effect or reaction of the material to the encountered tissue metabolism is also substantial [2]. The stabilization of the material in intra-osseous implantation or bone augmentation must be obtained [1]. To improve the stabilization, constituting the roughened surface on the material is one of the approaches [18]. In our study, the woven coating was considered a roughening method. Since increasing bone volume supports the biocompatibility of the material [12], the influence of the woven coating on bone reaction parameters was evaluated. Although the compositions of the FRC and FRC-C were similar, their ability to grow bone was uncertain. The biocompatibility of these materials was evaluated using osteoblastic or osteoclastic activities, periosteal reaction, and fibrosis, which indicate bone growth. According to our data, the first null hypothesis, that there would be no difference between the biocompatibilities of FRC and woven-coated FRC, was accepted. Previous in vivo studies demonstrated FRC causes neither inflammatory nor toxic reactions when encountering bone [8,9,16,21]. Besides, proper host response to FRC was observed in the osseous environment [12]. The results of our study were consistent with the literature. New bone formation around FRC and FRC-C specimens was observed at the 4th week of implantation and the formation of new bone was similar in FRC and woven-coated FRC; likewise, SEM investigation indicated a similar bone contact area of FRC and woven-coated FRC at the 12th week.
When considering the new bone formation of FRC, at the 4th week, the rate was 10%, while at the 12th week this rate was 50%. For the woven-coated FRC, these rates were 37.5% and 70%, respectively. This result can be explained by the fact that the maturation of the bone occurs and surrounds the implant over time. A previous study performed on the frontal bones of rabbits demonstrated the woven bone predominantly at the 3rd week, and the lamellar bone at the 6th and 8th weeks [12]. The higher rates of the contact area observed in woven-coated FRC can be explained by the high activity of osteoblasts due to the roughness. There is a certain relation between roughness and cell behavior. It demonstrated that the roughness of the surface increases osteoconduction via the promotion of differentiation and migration of the cells [22]. Much as the best scenario for bone ingrowth is approximately 1-2 μm of Ra value, with the 100-μm pore Fig. 1 The quality of bone union scores of FRC, FRC-C, and control group at the 4th and 12th weeks. FRC, fiberreinforced composite; FRC-C, woven-coated fiber-reinforced composite. Asterisk means statistical difference according to chi-square test (p<0.05) size, the needed situation for bone ingrowth and cell affinity was met according to the data of the literature [18,23].
In our study, we used woven laminate-coated FRC body and investigated biocompatibility. The contact of the woven laminates to bone showed new forming bone with no inflammation; this meant good biocompatibility. Another previous study reported the mature lamellar bone around BAG-coated FRC at the 12th week [13]. The mature bone was detected around the woven glass fiber-coated FRC at the 12th week, and this new bone-forming process was congruent with previous studies. However, to the best of our knowledge, there is no available data about the biocompatibility of the woven glass fiber-coated FRC body. The biocompatibility results of our in vivo study were similar to the biocompatibility of BAG-coated FRC. According to the SEM examinations, our study showed 85.5% and 86.3% of the bone contact area in FRC and woven-coated FRC, respectively. Thus, the second null hypothesis, that there would be a higher bone contact area in woven-coated FRC compared to FRC, was rejected. In the literature, previous data included the bone contact area of FRC and BAG-coated FRC that were implanted in the pig tibias at the 12th week which were 47% and 40%, Fig. 2 The scores of osteoblastic activity, osteoclastic activity, soft and hard tissue inflammation, periosteal reaction, and fibrosis of FRC, FRC-C, and control group at the 4th and 12th weeks. FRC, fiber-reinforced composite; FRC-C, woven-coated fiber-reinforced composite. Asterisk means statistical difference according to chisquare test (p<0.05)

Table 2
The scores (%) of bone reactions of groups in the 4th and 12th weeks. respectively [8]. Contrary to this result, in our study, the bone contact area did not reduce with the woven glass fiber coating. Besides, the different rates can be attributed to the differences in methodologies, animals, and implantation volume. With regard to the parameter of periosteal reaction, the results of woven-coated FRC were comparable with the control group. According to this, it can be thought that the rough structure of woven glass fiber promoted transversal growth. Bone metabolism is a process that included formation and resorption and depends on different factors. The osseointegration of the implant material is crucial for survival, but also, its behavior must be similar to the bone, which prevents improper stress distribution of the implant under mechanical loading and hence resorption [2,8,9]. The elastic modulus of FRC that is similar to bone enables proper strain in the complex that includes bone and implant.
Depending on the fact that proper stress transfer requires the concordance between the elastic modulus of different entities, its use for the reconstruction of maxillofacial defects can be considered compared to more rigid titanium material [1,2,16]. Dental surgery needs optimum material for the reconstruction of irregular bone defects that require esthetic and function at the same time. The elastic structure of FRC allows it adaptable to irregular bone defects. The woven laminate can be malleable before polymerization, and according to defect size and shape it can be shapeable. In addition, the reconstruction of a maxillofacial defect with polymer-based woven laminates would cause no radiographic artifacts that hamper the examination of all anatomic structures [3,8].
There was no sign of inflammation of the FRC and FRC-C which means the good biocompatibility of the material. On the other hand, the thick and macrophage or multinucleated giant cell-rich structure of the fibrous capsule can be attributed to the incompatibility of the implant material [3]. Our study showed that fibrous capsule was observed around the specimens. When the macrophages and multinuclear giant cells failed to disintegrate the foreign body by phagocytosis mechanism, the best scenario for the organism was to encapsulate the foreign body. The fibrous capsule can be explained by this mechanism. Another previous study reported a similar result that demonstrated a thicker fibrous capsule in FRC compared to titanium [3,16]. This response mechanism may be the best option for the host; however, the osseointegration of the implant material Titanium implants are still the most common treatment alternative due to their biomechanical properties [2]. Nevertheless, polymer-based implants including FRC have the potential depending on their proper elastic modulus compared to metallic ones. Besides, BAG-coated implants were developed to enhance the osseointegration of polymer-based implants [19]. The most important drawback of FRC that has a polymerization-based structure is the presence of residual monomers. The residual monomers may interfere with the biocompatibility of the dental implant materials. Besides, the double bonds that do not react result in the matrix being more prone to degradation [24]. However, our results showed good biocompatibility of FRC. This can be explained by the total polymerization of polymers. This result is congruent with a previous study that showed the well-polymerized material did not have inadequate biocompatibility [1]. FRC implants must have a maximum degree of polymerization to eliminate the toxicity related to the residual monomer.
The length and orientation of the fibers are crucial parameters that affect the mechanical properties of the material. FRCs can consist of fibers in continuous or discontinuous nature; besides, fiber direction can be unidirectional or bidirectional [25]. We used FRC which has continuous unidirectional fibers and FRC-C which has bidirectional mesh fibers in our study. The strength of continuous fibers is related to fiber orientation due to their anisotropic nature. Unidirectional orientation shows high endurance, specifically when the force is applied along the fiber direction [26]. This design may be advantageous in the case of the forces in one direction; however, it can remain incapable in the systems under multidirectional forces including maxillofacial defects or dental implants; thus, the graft must be designed carefully. Further studies are needed to investigate The limitations of this study are examination duration and limited information on humans due to performed on animals. We observed the biocompatibility and the response of vital tissues for up to 12th weeks and obtained no information on long-term outcomes. Further laboratory or clinical studies are needed to understand the long-term mechanical and biological effects of woven-coated FRC. Also, we examined the healing of bone in a limited volume; however, in clinical conditions there are bigger defects encountered. Therefore, further studies that investigate the bone response of higher volumes are also needed. While the woven coating did not negatively affect the biocompatibility, it did not also improve the bone contact area; therefore, fiber coatings with a different improved design that will increase osseointegration can be examined in further studies. As a consequence, it can be considered a potential material for a dental implant, reconstruction, or augmentation of the maxillofacial region depending on the promising results regarding the biological properties of FRC.

Conclusion
Within the limitation of this study, woven coating FRC demonstrated similar biocompatibility compared to FRC and showed no inflammation. On the other hand, it did not enhance the quality of bone union, osteoblast activity, periosteal reaction, and bone contact area. Its potential use as a dental implant, reconstruction material, and scaffold for augmentation can be considered.