Materials Science and Engineering C 36 (2014) 294–300
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Biodegradable borosilicate bioactive glass scaffolds with a trabecular microstructure for bone repair Yifei Gu a, Gang Wang b, Xin Zhang a, Yadong Zhang b, Changqing Zhang b, Xin Liu c, Mohamed N. Rahaman c, Wenhai Huang a,⁎, Haobo Pan d a
Department of Materials Science and Engineering, Tongji University, Shanghai 200092, China Department of Orthopedic Surgery, Shanghai Sixth People's Hospital, Shanghai Jiao Tong University, Shanghai 200233, China Department of Materials Science and Engineering, and Center for Bone and Tissue Repair and Regeneration, Missouri University of Science and Technology, Rolla, MO 65409-0340, USA d Department of Orthopaedics & Traumatology, The University of Hong Kong, 999077, Hong Kong, China b c
a r t i c l e
i n f o
Article history: Received 29 January 2013 Received in revised form 25 October 2013 Accepted 17 December 2013 Available online 27 December 2013 Keywords: Bioactive glass Scaffold Polymer foam replication In vitro degradation Bone regeneration Platelet-rich plasma
a b s t r a c t Three-dimensional porous scaffolds of a borosilicate bioactive glass (designated 13-93B1), with the composition 6Na2O–8K2O–8MgO–22CaO–18B2O3–36SiO2–2P2O5 (mol%), were prepared using a foam replication technique and evaluated in vitro and in vivo. Immersion of the scaffolds for 30 days in a simulated body fluid in vitro resulted in partial conversion of the glass to a porous hydroxyapatite composed of fine needle-like particles. The capacity of the scaffolds to support bone formation in vivo was evaluated in non-critical sized defects created in the femoral head of rabbits. Eight weeks post-implantation, the scaffolds were partially converted to hydroxyapatite, and they were well integrated with newly-formed bone. When loaded with platelet-rich plasma (PRP), the scaffolds supported bone regeneration in segmental defects in the diaphysis of rabbit radii. The results indicate that these 13-93B1 scaffolds, loaded with PRP or without PRP, are beneficial for bone repair due to their biocompatibility, conversion to hydroxyapatite, and in vivo bone regenerative properties. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Large bone defects resulting from trauma, resection for tumors, and congenital diseases are a common clinical problem. Whereas microdefects in bone can heal over time, large defects are difficult to repair without external intervention such as bone grafting. At present, bone autograft, bone allograft, and prosthetic implants are widely used to repair large bone defects, but they suffer from limitations. Autologous bone grafts suffer from problems such as limited availability and donor site morbidity, whereas bone allografts and prosthetic implants show uncertain healing to bone, unpredictable long-term durability, and also suffer from high costs. Porous synthetic scaffolds that mimic bone would be ideal bone substitutes, but such porous scaffolds should have the capacity to support tissue ingrowth and integration with host bone and surrounding soft tissues, and they should degrade at a rate compatible with new bone formation. Bioactive glasses have several attractive properties as a scaffold material in bone repair. Apart from being biocompatible, bioactive glasses degrade and convert to hydroxyapatite (HA) in vivo, which promotes osseous healing [1]. Calcium ions and soluble silicon released during the conversion of silicate bioactive glass such as 45S5 further promote osteogenesis [2,3]. Bioactive glasses are also amenable to fabrication ⁎ Corresponding author. Tel./fax: +86 21 65980040. E-mail addresses: [email protected], [email protected] (W. Huang). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.12.023
into porous three-dimensional (3D) architectures that are capable of supporting tissue ingrowth and integration. Our previous work has shown that by replacing varying amounts of SiO2 in silicate 45S5 or 13-93 glass (6Na2O–8K2O–8MgO–22CaO– 54SiO2–2P2O5; mol%) with B2O3, borosilicate and borate bioactive glasses with a controllable degradation rate can be produced [4,5]. In particular, the degradation rate of the glass and its conversion to HA increase with the replacement of higher amounts of SiO2 in silicate bioactive glass with B2O3. A borate glass (designated 13-93B3) with the composition 6Na2O–8K2O–8MgO–22CaO–54B2O3–2P2O5 (mol%), obtained by replacing all the SiO2 in 13-93 glass with B2O3, was shown to convert completely to HA when immersed in an aqueous phosphate solution, at a rate that was ~3–4 times faster than 13-93 bioactive glass [4,5]. However, the rapid degradation in strength which results from the conversion process limits the application of borate 13-93B3 scaffolds to the repair of non-loaded bone defects. Borate 13-93B3 glass has been successfully used as a drug delivery system for vancomycin [6,7] and teicoplanin [8,9] in the treatment of osteomyelitis in a rabbit tibial model. In addition, bone regeneration was observed in the sites implanted with the borate glass carrier but not with a carrier composed of commercial calcium sulfate beads [10]. A borosilicate glass (designated 13-93B1) with the composition 6Na2O–8K2O–8MgO–22CaO–18B2O3–36SiO2–2P2O5 (mol%), obtained by replacing one-third of the molar concentration of SiO2 in 13-93 glass with B2O3, was observed to degrade and convert faster to HA
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
than silicate 13-93 glass but slower than borate 13-93B3 [4,5]. Consequently, scaffolds of 13-93B1 glass could retain a greater fraction of their strength over a longer period of time in an aqueous phosphate solution when compared to borate 13-93B3 scaffolds. Borosilicate 1393B1 glass also showed the capacity to support the proliferation and function of osteogenic MLO-A5 cells in vitro [11]. When implanted subcutaneously in the dorsum of rats for 6 weeks, porous 13-93B1 scaffolds showed the capacity to support tissue infiltration into the interior pores [11]. Based on the promising in vitro and in vivo results described above, this study was undertaken to evaluate the capacity of 13-93B1 glass scaffolds to repair bone defects in vivo. Scaffolds with a “trabecular” microstructure, similar to that of dry human trabecular bone, were prepared using a foam replication technique. The conversion of the glass to HA was studied in simulated body fluid (SBF) in vitro. The capacity of the scaffolds to support bone regeneration in vivo was evaluated using a rabbit femoral defect model. Since growth factors are often required to stimulate bone formation [12], the capacity of the scaffolds to serve as a carrier for platelet-rich plasma (PRP) was also evaluated in segmental defects created in the diaphysis of rabbit radii. 2. Materials and methods 2.1. Preparation of borosilicate (13-93B1) bioactive glass scaffolds Scaffolds of borosilicate (13-93B1) bioactive glass scaffolds were prepared using a polymer foam replication technique, as described previously [13]. This technique was used because of its ability to produce scaffolds with a microstructure similar to dry human trabecular bone. Briefly, glass frits with the composition 6Na2O–8K2O–8MgO–22CaO– 18B2O3–36SiO2–2P2O5 (mol%) were prepared by melting the required quantities of Na2CO3, K2CO3, MgCO3, CaCO3, H3BO3, SiO2, and NaH2PO4 · 2H2O (analytical grade; Sinopharm Chemical Reagent Co., Ltd., China) in a platinum crucible for 30 min at 1150 °C in air, and quenching the melt in cold water. Glass particles of size b 50 μm were obtained by crushing the glass frits with a hardened steel mortar and pestle and sieving through a stainless steel sieve. A slurry for the polymer foam infiltration step was prepared by dispersing the 13-93B1 particles in ethanol using ethyl-cellulose (EC; analytical grade, Sinopharm Chemical Reagent Co., Ltd., China) as a dispersant and binder. Cylindrical samples (5 mm in diameter × 15 mm) of a polyurethane foam (Shanghai No. 6 Plastic Co., Ltd., China), with open porosity of ~ 50 pores per inch, were immersed in the slurry to coat them with a layer of glass particles. After drying overnight at room temperature, the coated foams were heated in air for 1 h at 500 °C (heating rate = 1 °C/min) to burn off the foam and polymeric additives in the glass coating, and then for 2 h at 560 °C (heating rate = 5 °C/min) to sinter the glass particles into a dense network.
295
2.3. Evaluation of the bioactivity of 13-93B1 glass in vitro The bioactivity of the 13-93B1 glass was assessed in vitro from the conversion of as-fabricated scaffolds and disks (10 mm in diameter × 2 mm) in simulated body fluid (SBF) at 37 °C. One gram of glass was immersed in 100 ml SBF, as described in previous studies [5,14,15], and immersion times of up to 30 days were used. At selected times, the scaffolds and disks were removed from the SBF, washed twice with deionized water and then twice with ethanol, and dried at 90 °C. The weight loss of the scaffolds and disks was determined as ΔW = (Wo − Wt) / Wo, where Wo is the initial mass of the sample, and Wt is the mass after immersion for time t in the SBF. Four replicates were used for each time point, and the weight loss was determined as a mean ± standard deviation (SD). After the samples were removed at each time point, the SBF was cooled to room temperature and its pH was measured. The crystalline phase formed on the surface of the glass disks by conversion in SBF was analyzed using thin-film XRD (X'Pert Pro; PANalytical, Almelo, The Netherlands) using Ca Kα1 radiation (λ = 0.15405 nm; incident radiation = 45 kV) at a scanning rate of 0.03°/min in the 2θ range 10–80°C. The surface morphology and microstructure of the scaffolds after immersion in SBF were examined in a field-emission SEM (Quanta 200 FEG) using the conditions described earlier. 2.4. Evaluation of scaffolds in rabbit femoral head defect model Three male New Zealand white (NZW) rabbits weighing 2.5–3.0 kg were used in the experiments. The animals were obtained from the Shanghai Laboratory Animal Center (Shanghai, China, Certificate number SCXK: 2002-0010). Animal care and surgical procedures were in accordance with guidelines issued by the Department of Science and Technology of China in 2006 (Guidance Suggestions from the Act for Care and Use of Laboratory Animals). Under intramuscular anesthesia (3% pentobarbital sodium; 30 mg per kg mass of rabbit), lateral approaches were performed in both shaved front knees to expose the distal femoral diaphysis. One defect (5 mm diameter × 5 mm deep) was created in each of those femoral heads using a medium speed burr under constant irrigation with sterile saline. The defects were implanted with the scaffolds or left unfilled (control), and the wounds were sutured. At 4 and 8 weeks post-implantation, the femoral heads with the implants or unfilled defects were harvested and fixed in 10% neutral formalin–saline solution. The samples were sectioned in the center to expose the cross section of the implants, and decalcified in 10% formic acid. After the samples were dehydrated in a graded series of alcohol, rinsed in xylene, and embedded in paraffin, 5 μm thick sections were cut and stained with hematoxylin and eosin (H&E). The stained sections were examined in an optical microscope (AX80T, Olympus, Japan).
2.2. Characterization of 13-93B1 scaffolds 2.5. Evaluation of scaffolds in critical-sized segmental defects in rabbit radii The as-fabricated scaffolds were coated with Au/Pd and their microstructure was examined in a field-emission scanning electron microscope, SEM (Quanta 200 FEG; FEI Co., The Netherlands) at an accelerating voltage of 10 kV and a working distance of 7.5 mm. The porosity in the scaffold was measured using the Archimedes method, while the size distribution of the open pores was measured using a liquid extrusion porosimeter (LEP-1100 AX, Porous Materials Inc., NY) with water as the wetting liquid. According to the manufacturer's instruction, the distribution of pore size is defined as [−(dV / d log D)], where V is the pore volume in 1 g porous scaffold and D is the pore diameter. The compressive strength of cylindrical samples (6 mm in diameter × 12 mm) was measured using an Instron testing machine (Model 4881; Instron Co., Norwood, MA) at a crosshead speed of 0.5 mm/min. Five samples were tested, and the strength was expressed as a mean ± standard deviation.
Two groups of 13-93B1 scaffolds were evaluated in critical-sized segmental defects created in the diaphysis of rabbit radii: (1) asfabricated 13-93B1 scaffolds and (2) as-fabricated 13-93B1 scaffolds loaded with platelet-rich plasma (PRP). The unfilled defects served as the negative control group. Loading the scaffolds with PRP was performed as follows. Rabbits to be implanted with the PRP-loaded scaffolds were anesthetized by intravenously injecting 25 g/l pentobarbital sodium at a dose of 1 ml per kg body weight. Then 5 ml of blood was aspirated from the central arteries of the rabbit ears. The platelets in the blood were enriched by a two-step centrifugation process. First, the erythrocytes were removed at 8.40 N (centrifugal force at 2500 rev/min; 10 min) in a centrifuge (Heraeus Labofuge 300, Kendro Lab Products, Langenselbold, Germany) at 20 °C. Second, the leukocytes were sedimented at 13.10 N (centrifugal
296
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
3. Results 3.1. Characteristics of as-fabricated 13-93B1 scaffolds The microstructure of the as-fabricated 13-93B1 glass scaffolds (Fig. 1a) was similar to that of dry human trabecular bone. The scaffolds had a porosity of 78 ± 8% and a pore size in the range 400–650 μm, with an average of ~500 μm (Fig. 1b). The compressive strength of the as-fabricated scaffolds was 5.1 ± 1.7 MPa. 3.2. Degradation and conversion of 13-93B1 glass in vitro Immersion of the as-fabricated scaffolds in SBF resulted in the degradation and conversion of the smooth glass surface (Fig. 1a) to a porous layer composed of fine needle-like particles (Fig. 2a, b). Thin-film XRD of 13-93B1 glass disks immersed in SBF for 30 days showed diffraction peaks that corresponded to those of a reference HA (JCPDS 09-0432), confirming that the converted material was HA (Fig. 3). The broad peaks with a low intensity indicated that the HA was poorly crystallized or was composed of nanocrystalline particles. Fig. 4 shows the weight loss of the scaffolds and the pH of the SBF as a function of immersion time. The trend in the data, consisting of a rapid increase initially followed by a slower increase thereafter, is typically observed for the conversion of silicate, borosilicate, and borate bioactive glasses to HA [4,5]. After 30 days of immersion, the weight loss of the scaffolds was ~35%, while the pH of the SBF increased from an initial value of 7.4 to ~8.2. 3.3. Bone regeneration in rabbit femoral defects No post-operative complications or inflammation were observed at 4 or 8 weeks post-implantation. Fig. 5 shows the gross appearance of
Fig. 1. (a) SEM image of 13-93B1 scaffold prepared by a foam replication technique; (b) pore size distribution of the fabricated scaffold.
force at 4000 rev/min; 10 min) at 20 °C to obtain PRP with a platelet concentration equal to ~5.5 times the value for normal blood. A volume of 0.8 ml PRP was pipetted onto each scaffold and gelled by adding 0.2 ml thrombin before implantation. Eighteen NZW rabbits were used in these experiments. The rabbits were anesthetized by intravenously injecting 0.5 ml per kg body weight of ketamine (42.8 mg/ml) and xylazine (0.7 mg/ml). The rabbits were immobilized on their backs and the right hind limbs were shaved and disinfected with povidone–iodine. A longitudinal incision of 2.5–3 cm in the skin was made over the distal radius. Then the extensor retinaculum was separated to expose the bone. A segmental diaphyseal defect (1.5-cm long) was created with an oscillating saw under irrigation with sterile saline. The defects were implanted with 6 implants per group (as-fabricated scaffolds; as-fabricated scaffolds with PRP), while 6 unfilled defects served as the control group. At 4 or 8 weeks post-implantation, the rabbits were sacrificed by intravenous injection with an overdose of pentobarbital sodium. 2.6. Radiography and histology X-ray radiographs were taken of the right tibiae of the rabbits. After the removal of surrounding soft tissue and bone, the implants were photographed using a digital camera. The harvested implants were fixed in 10% formalin−saline solution and decalcified in 10% formic acid. After the samples were dehydrated in a graded series of alcohol and embedded in paraffin, longitudinal sections (5 μm thick) were cut and stained with H&E. The stained sections were examined in a transmitted light microscope (AX80T, Olympus).
Fig. 2. SEM images of the surface of 13-93B1 scaffold after immersion for 30 days in simulated body fluid (SBF): (a) lower magnification; (b) higher magnification.
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
297
Fig. 6 shows H&E stained sections of defects implanted with the 1393B1 glass scaffolds for 4 and 8 weeks. The unconverted glass was shown in white due to the decalcification, while the converted HA was stained in pink because of the adsorption of intracellular or extracellular proteins [7], indicative of partial conversion of the glass to HA. At 4 weeks, soft tissues grew into the HA formed by degradation and conversion of the glass, while new bone (denoted NB) infiltrated the macropores of the scaffold (denoted S). The amount of new bone increased at 8 weeks, and the new bone was well integrated with the HA that formed on the glass. A larger number of osteoblast-like cells were apparent in the tissues at 8 weeks than at 4 weeks.
3.4. Repair of segmental defects in rabbit radii
Fig. 3. (a) Thin-film XRD pattern of the surface of 13-93B1 glass disk after immersion for 30 days in SBF; (b) XRD pattern of a reference hydroxyapatite (JCPDS 09-0432).
Fig. 4. Weight loss of 13-93B1 scaffolds and pH of the solution as a function of immersion time of the scaffolds in SBF at 37 °C.
the defects implanted with the 13-93B1 scaffolds for 4 and 8 weeks and the unfilled defects (control) at 8 weeks. The defects implanted with the scaffolds showed better bone healing than the unfilled control group. Bone healing improved with implantation time for the defects implanted with the scaffolds (Fig. 5a, b). In comparison, the unfilled defects showed little new tissue formation, and the defects were still clearly evident. Though the defects were not created in the exact same position, they were still in the femoral head area. Bone response in this area could be viewed the same. However, the technique to create the defect in the femoral head should be improved in our future work.
After surgery, the animals implanted with the 13-93B1 scaffolds (without or with PRP) showed no signs of inflammation or infection, except for one animal implanted with a scaffold without PRP which showed symptoms of slight inflammation. Fig. 7 shows the appearance of segmental defects in the rabbit radii implanted with the two groups of scaffolds and the unfilled defects at 8 weeks. Both groups of scaffolds integrated with host bone, and callus formation apparently originated from both ends of the defect (Fig. 7a, b). In comparison, no bone formation was apparent in the unfilled defect (Fig. 7c). Radiographic images of the implants at 4 weeks showed a distinct radiolucent zone at the interface between the implant and the host bone (Fig. 8a, b), confirming the integration of the implants with host bone. For both groups of scaffolds, an osseous callus also formed around the implant at 4 weeks. At 8 weeks (Fig. 8a, b), the presence of the osseous callus was more evident in the defects implanted with the PRPloaded scaffolds. The radiographs also showed a decrease in the density of the material in the defect sites, particularly for the PRP-loaded implants, and the ends of the implants were integrated with host bone. The osseous callus grew faster and was more evident in the defects implanted with the PRP-loaded scaffolds than in the defects implanted with the scaffolds without PRP. For the unfilled defects, a small amount of osseous callus was observed at the edges of the defect and on the side of the ulna at 8 weeks, but the edges of the defect were still clearly visible (Fig. 8c). The H&E stained images of the defects showed that bone formation increased during the eight-week implantation period for both groups of scaffolds (with or without PRP) (Fig. 9), an observation that was consistent with the radiographic analysis described above. However, there were also differences in bone regeneration between the two groups of scaffolds. Four weeks post-implantation, while the amount of new bone (NB) appeared to be small, there appeared to be a greater amount of new bone formed in the defects implanted with the PRP-loaded scaffolds. The defects contained numerous osteoblast-like cells (arrows) that penetrated around and within the pores of the scaffold (S). Woven bone, as well as cancellous and lamellar bone (LB) were observed, particularly in the scaffolds loaded with PRP.
Fig. 5. Gross appearance of rabbit femoral head defects implanted with 13-93B1 scaffolds for (a) 4 weeks, and (b) 8 weeks; (c) unfilled defect at 8 weeks.
298
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
Fig. 6. H&E stained sections of femoral head defects implanted with 13-93B1 scaffolds for (a) 4 weeks, and (b) 8 weeks. S: scaffold; NB: new bone.
4. Discussion The architecture of the scaffold is known to affect tissue ingrowth, cell attachment and diffusion of nutrients into the defect site. Studies have shown that interconnected pores of size N100 μm are required to support bone ingrowth [16], but an increase in the size of the interconnected pores of the scaffold can lead to increased bone growth by allowing a greater migration of mesenchymal cells, osteoblasts, and neovascularization [17]. The borosilicate bioactive glass (13-93B1) used in the present study was fabricated into a “trabecular” microstructure with a porosity of 78 ± 8%, and a pore size of 400–650 μm, providing pore characteristics that are known to be favorable for bone ingrowth.
When immersed in SBF, the glass reacted with the solution to form a surface layer of HA (Figs. 2, 3), which is an indication of its bioactivity in vitro and its bioactive potential in vivo [14]. It was reported that the release of calcium and phosphate ions from the implants and subsequent formation of a biological HA layer could promote bone formation in vivo [18]. The adsorption of proteins from serum or the body fluid onto the high-surface-area HA layer is also reported to be important for new bone formation [19]. The degradation and conversion of the glass to HA resulted in a weight loss of the scaffolds and an increase in the pH of the SBF (Fig. 4). Assuming that all the CaO in the glass reacted to form HA with a stoichiometric composition, the theoretical weight loss of the fully converted 13-93B1 glass is ~65% [15]. After 30 days of immersion in SBF, the measured weight loss was ~ 35%, which indicates that the glass was only partially converted to HA. However, it is known that the conversion of 13-93B1 glass is faster than the slowly-degrading silicate 13-93 glass but also much slower than borate 13-93B3 glass [5,15]. Consequently, the use of 13-93B1 glass as a scaffold material could provide an optimal combination of conversion to HA and strength retention which cannot be readily achieved with 13-93 or 13-93B3 glass. The increase in the pH of the SBF during the conversion process is attributed to the dissolution of alkali ions (Na+ and K+) from the glass and the consumption of phosphate ions from the SBF. Since phosphoric acid is a stronger acid than silicic or boric acid, dissolution of Si and B from the glass, presumably as silicic acid and borate ions, respectively, coupled with the consumption of phosphate ions could also contribute to the pH increase. However, the amount of ions released in SBF, including Ca2+, Na+ and K+, was not measured in the present work. Scaffolds for bone repair should also have the requisite mechanical properties to support physiological loads and they should degrade in tune with new bone formation. As the scaffold degrades and loses strength [15], new bone formed within the pores of the scaffold should act to offset the decrease in strength. The compressive strength of the as-fabricated 13-93B1 scaffolds (5.1 ± 1.7 MPa) was within the range of strengths (2–12 MPa) reported for human trabecular bone. Consequently, the bioactive glass scaffolds used in this study could potentially be applied as a bone grafting material in the regeneration of non-loaded bone. A concern with scaffolds composed of borosilicate and, in particular, borate bioactive glasses is the toxicity of boron released during the degradation and conversion of the scaffolds to HA. While low concentrations of boron are considered to be beneficial for the formation of healthy bone [20,21], high boron concentrations can result in chronic toxicity [22,23]. Our previous work showed that when the amount of boron released is below a certain threshold concentration, boratebased bioactive glasses can support the proliferation of bone marrow stromal cells and MLO-A5 cells [24]. Bone regeneration is also largely dependent on the degradation of the scaffolds at an optimal rate and the formation of non-cytotoxic degradation products [25,26].
Fig. 7. Gross appearance of segmental defect sites in rabbit radii after implantation for 8 weeks with (a) 13-93B1 scaffolds, and (b) 13-93B1 scaffolds loaded with platelet-rich plasma; (c) unfilled defect at 8 weeks. S: scaffold; OC: osseous callus.
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
299
Fig. 8. X-ray radiographs of rabbit radius defect sites after implantation for 4 and 8 weeks with (a) 13-93B1 scaffolds, and (b) 13-93B1 scaffolds loaded with platelet-rich plasma; (c) unfilled defect at 4 and 8 weeks.
Two osseous defect models were used in this study to assess the performance of the 13-93B1 scaffolds in vivo. The rabbit model is commonly used for screening implant materials prior to testing in a larger animal model [2,27–29]. The non-critical-sized defect in a rabbit femoral head model has been used extensively in previous studies to rapidly assess bone healing, whereas a segmental defect in rabbit radii was used as a critical-sized defect model. The results showed that both groups of scaffolds were biocompatible and supported the formation of new bone in the osseous defects mainly by osteoconduction. However, scaffolds loaded with PRP appeared to show a greater capacity to support bone formation and healing of the defects when compared to the scaffolds without PRP. Several growth factors are present in PRP, such as transforming growth factor β1, insulin-like growth factor, and vascular endothelial growth factor, which are reported to be beneficial for stimulating bone formation. While there are conflicting reports on the role of PRP in enhancing bone healing [30–32], the greater ability of the PRP-loaded scaffolds in enhancing bone formation observed in this study confirmed the positive effect of PRP on bone healing. Presumably the growth factors provided the PRP-loaded scaffolds with an osteoinductive property, in addition to the inherent osteoconductive property of the porous scaffolds themselves, by releasing the growth factors to stimulate the
proliferation and differentiation of osteoblastic cells. Because PRP has a limited stability and a short life span, its effectiveness in bone healing is dependent on the use of a suitable carrier to provide a controlled release over the required time period of PRP stability. Based on the ability of the PRP-loaded scaffolds to enhance bone regeneration, the porous 13-93B1 scaffolds used in the present study apparently provided an effective carrier. In general, the results of the present study indicated that loading bioactive glass scaffolds with PRP could provide a promising approach for bone regeneration. However further experiments are needed to better control the release rate of the growth factors from the scaffold and the degradation rate of the bioactive glass scaffold. 5. Conclusion Borosilicate (13-93B1) bioactive glass scaffolds prepared by a polymer foam replication technique had pore characteristics (porosity = 78%; pore size = 400–650 μm) that are known to be favorable for supporting bone ingrowth. The scaffolds degraded and converted to hydroxyapatite when immersed in a simulated body fluid in vitro, indicating their bioactive potential. The scaffolds supported bone regeneration when implanted for 8 weeks in non-critical-sized defects in the femoral head of rabbits or in critical-sized segmental
Fig. 9. H&E stained sections of rabbit radius defects after implantation for 4 and 8 weeks with (a) 13-93B1 scaffolds, and (b) 13-93B1 scaffolds loaded with platelet-rich plasma; S: scaffold; NB: new bone; LB: lamellar bone. Arrows indicate osteocytes.
300
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
defects in rabbit radii. Loading the 13-93B1 scaffolds with platelet-rich plasma enhanced bone regeneration in the segmental defects when compared to the 13-93B1 scaffolds without platelet-rich plasma. The combination of borosilicate 13-93B1 bioactive glass scaffolds and growth factors could provide a promising approach in bone repair. Acknowledgment This research was supported by the Shanghai Committee of Science and Technology (Grant No. 12JC1408500) and the National Natural Science Foundation (Grant Nos. 51072133, 51372170 & 51272274). References [1] L.L. Hench, J. Am, Ceram. Soc. 74 (1991) 1487–1510. [2] D.L. Wheeler, K.E. Stokes, H.M. Park, J.O. Hollinger, J. Biomed. Mater. Res. 35 (1997) 249–254. [3] D.L. Wheeler, K.E. Stokes, R.G. Hoellrich, D.L. Chamberland, S.W. McLoughlin, J. Biomed. Mater. Res. 41 (1998) 527–533. [4] W. Huang, D.E. Day, K. Kittiratanapiboon, M.N. Rahaman, J. Mater, Sci.: Mater. Med. 17 (2006) 583–596. [5] A. Yao, D. Wang, W. Huang, Q. Fu, M.N. Rahaman, D.E. Day, J. Am. Ceram. Soc. 90 (2007) 303–306. [6] Z. Xie, X. Liu, W. Jia, C. Zhang, W. Huang, J. Wang, J. Control. Release 139 (2009) 118–126. [7] X. Liu, Z. Xie, C. Zhang, H. Pan, M.N. Rahaman, X. Zhang, Q. Fu, W. Huang, J. Mater. Sci. Mater. Med. 21 (2009) 575–582. [8] X. Zhang, W. Jia, Y. Gu, W. Xiao, X. Liu, D. Wang, C. Zhang, W. Huang, M.N. Rahaman, D.E. Day, N. Zhou, Biomaterials 31 (2010) 5865–5874. [9] W.-T. Jia, X. Zhang, S.-H. Luo, X. Liu, W.-H. Huang, M.N. Rahaman, D.E. Day, C.-Q. Zhang, Z.-P. Xie, J.-Q. Wang, Acta Biomater. 6 (2010) 812–819. [10] Q. Fu, W. Huang, W. Jia, M.N. Rahaman, X. Liu, A.P. Tomsia, Tissue Eng. Part A 17 (2011) 3077–3084.
[11] Q. Fu, M.N. Rahaman, B.S. Bal, L.F. Bonewald, K. Kuroki, R.F. Brown, J. Biomed, Mater. Res. Part A 95A (2010) 172–179. [12] D.F. Williams, Biomaterials 30 (2009) 5897–5909. [13] X. Liu, A. Yao, D. Wang, W. Huang, H. Fu, D. Zhao, W. Lu, H. Pan, J. Inorg. Mater. 23 (2008) 322–326. [14] D. Zhao, W. Huang, M.N. Rahaman, D.E. Day, D. Wang, Acta Biomater. 5 (2009) 1265–1273. [15] Q. Fu, M.N. Rahaman, H. Fu, X. Liu, J. Biomed, Mater. Res. Part A 95A (2010) 164–171. [16] S.F. Hulbert, F.A. Young, R.S. Mathews, J.J. Klawitter, C.D. Talbert, F.H. Stelling, J. Biomed. Mater. Res. 4 (1970) 433–456. [17] V. Karageorgiou, D. Kaplan, Biomaterials 26 (2005) 5474–5491. [18] T. Hanawa, Y. Kamiura, S. Yamamoto, T. Kohgo, A. Amemiya, H. Ukai, K. Murakami, K. Asaoka, J. Biomed. Mater. Res. 36 (1997) 131–136. [19] X.D. Zhu, H.S. Fan, Y.M. Xiao, D.X. Li, H.J. Zhang, T. Luxbacher, X.D. Zhang, Acta Biomater. 5 (2009) 1311–1318. [20] F.H. Nielsen, Plant Soil 193 (1997) 199–208. [21] F.H. Nielsen, Nutr. Rev. 66 (2008) 183–191. [22] F.J. Murray, Regul. Toxicol. Pharmacol. 22 (1995) 221–230. [23] F.J. Murray, J. Trace Elem, Exp. Med. 9 (1996) 231–243. [24] H. Fu, Q. Fu, N. Zhou, W. Huang, M.N. Rahaman, D. Wang, X. Liu, Mater. Sci. Eng. C 29 (2009) 2275–2281. [25] E. Alsberg, H.J. Kong, Y. Hirano, M.K. Smith, A. Albeiruti, D.J. Mooney, J. Dent. Res. 82 (2003) 903–908. [26] H.-J. Sung, C. Meredith, C. Johnson, Z.S. Galis, Biomaterials 25 (2004) 5735–5742. [27] C.G. Trejo, D. Lozano, M. Manzano, J.C. Dodrio, A.J. Salinas, S. Dapía, E. Gómez-Barrena, M. Vallet-Regí, N. García-Honduvilla, J. Buján, P. Esbrit, Biomaterials 31 (2010) 8564–8573. [28] H. Liao, X.F. Walboomers, W.J.E.M. Habraken, Z. Zhang, Y. Li, D.W. Grijpma, A.G. Mikos, J.G.C. Wolke, J.A. Jansen, Acta Biomater. 7 (2011) 1752–1759. [29] P. Niemeyer, K. Szalay, R. Luginbühl, N.P. Südkamp, P. Kasten, Acta Biomater. 6 (2010) 900–908. [30] J. Iqbal, S.H. Pepkowitz, E. Klapper, Curr. Osteoporos. Rep. 9 (2011) 258–263. [31] A. Oryan, A. Meimandi Parizi, Z. Shafiei-Sarvestani, A.S. Bigham, Cell Tissue Banking 13 (4) (2012) 639–651. [32] R.E.C.M. Mooren, M.A.W. Merkx, E.M. Bronkhorst, J.A. Jansen, P.J.W. Stoelinga, Int. J. Oral Surg. 36 (2007) 626–631.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Biodegradable borosilicate bioactive glass scaffolds with a trabecular microstructure for bone repair Yifei Gu a, Gang Wang b, Xin Zhang a, Yadong Zhang b, Changqing Zhang b, Xin Liu c, Mohamed N. Rahaman c, Wenhai Huang a,⁎, Haobo Pan d a
Department of Materials Science and Engineering, Tongji University, Shanghai 200092, China Department of Orthopedic Surgery, Shanghai Sixth People's Hospital, Shanghai Jiao Tong University, Shanghai 200233, China Department of Materials Science and Engineering, and Center for Bone and Tissue Repair and Regeneration, Missouri University of Science and Technology, Rolla, MO 65409-0340, USA d Department of Orthopaedics & Traumatology, The University of Hong Kong, 999077, Hong Kong, China b c
a r t i c l e
i n f o
Article history: Received 29 January 2013 Received in revised form 25 October 2013 Accepted 17 December 2013 Available online 27 December 2013 Keywords: Bioactive glass Scaffold Polymer foam replication In vitro degradation Bone regeneration Platelet-rich plasma
a b s t r a c t Three-dimensional porous scaffolds of a borosilicate bioactive glass (designated 13-93B1), with the composition 6Na2O–8K2O–8MgO–22CaO–18B2O3–36SiO2–2P2O5 (mol%), were prepared using a foam replication technique and evaluated in vitro and in vivo. Immersion of the scaffolds for 30 days in a simulated body fluid in vitro resulted in partial conversion of the glass to a porous hydroxyapatite composed of fine needle-like particles. The capacity of the scaffolds to support bone formation in vivo was evaluated in non-critical sized defects created in the femoral head of rabbits. Eight weeks post-implantation, the scaffolds were partially converted to hydroxyapatite, and they were well integrated with newly-formed bone. When loaded with platelet-rich plasma (PRP), the scaffolds supported bone regeneration in segmental defects in the diaphysis of rabbit radii. The results indicate that these 13-93B1 scaffolds, loaded with PRP or without PRP, are beneficial for bone repair due to their biocompatibility, conversion to hydroxyapatite, and in vivo bone regenerative properties. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Large bone defects resulting from trauma, resection for tumors, and congenital diseases are a common clinical problem. Whereas microdefects in bone can heal over time, large defects are difficult to repair without external intervention such as bone grafting. At present, bone autograft, bone allograft, and prosthetic implants are widely used to repair large bone defects, but they suffer from limitations. Autologous bone grafts suffer from problems such as limited availability and donor site morbidity, whereas bone allografts and prosthetic implants show uncertain healing to bone, unpredictable long-term durability, and also suffer from high costs. Porous synthetic scaffolds that mimic bone would be ideal bone substitutes, but such porous scaffolds should have the capacity to support tissue ingrowth and integration with host bone and surrounding soft tissues, and they should degrade at a rate compatible with new bone formation. Bioactive glasses have several attractive properties as a scaffold material in bone repair. Apart from being biocompatible, bioactive glasses degrade and convert to hydroxyapatite (HA) in vivo, which promotes osseous healing [1]. Calcium ions and soluble silicon released during the conversion of silicate bioactive glass such as 45S5 further promote osteogenesis [2,3]. Bioactive glasses are also amenable to fabrication ⁎ Corresponding author. Tel./fax: +86 21 65980040. E-mail addresses: [email protected], [email protected] (W. Huang). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.12.023
into porous three-dimensional (3D) architectures that are capable of supporting tissue ingrowth and integration. Our previous work has shown that by replacing varying amounts of SiO2 in silicate 45S5 or 13-93 glass (6Na2O–8K2O–8MgO–22CaO– 54SiO2–2P2O5; mol%) with B2O3, borosilicate and borate bioactive glasses with a controllable degradation rate can be produced [4,5]. In particular, the degradation rate of the glass and its conversion to HA increase with the replacement of higher amounts of SiO2 in silicate bioactive glass with B2O3. A borate glass (designated 13-93B3) with the composition 6Na2O–8K2O–8MgO–22CaO–54B2O3–2P2O5 (mol%), obtained by replacing all the SiO2 in 13-93 glass with B2O3, was shown to convert completely to HA when immersed in an aqueous phosphate solution, at a rate that was ~3–4 times faster than 13-93 bioactive glass [4,5]. However, the rapid degradation in strength which results from the conversion process limits the application of borate 13-93B3 scaffolds to the repair of non-loaded bone defects. Borate 13-93B3 glass has been successfully used as a drug delivery system for vancomycin [6,7] and teicoplanin [8,9] in the treatment of osteomyelitis in a rabbit tibial model. In addition, bone regeneration was observed in the sites implanted with the borate glass carrier but not with a carrier composed of commercial calcium sulfate beads [10]. A borosilicate glass (designated 13-93B1) with the composition 6Na2O–8K2O–8MgO–22CaO–18B2O3–36SiO2–2P2O5 (mol%), obtained by replacing one-third of the molar concentration of SiO2 in 13-93 glass with B2O3, was observed to degrade and convert faster to HA
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
than silicate 13-93 glass but slower than borate 13-93B3 [4,5]. Consequently, scaffolds of 13-93B1 glass could retain a greater fraction of their strength over a longer period of time in an aqueous phosphate solution when compared to borate 13-93B3 scaffolds. Borosilicate 1393B1 glass also showed the capacity to support the proliferation and function of osteogenic MLO-A5 cells in vitro [11]. When implanted subcutaneously in the dorsum of rats for 6 weeks, porous 13-93B1 scaffolds showed the capacity to support tissue infiltration into the interior pores [11]. Based on the promising in vitro and in vivo results described above, this study was undertaken to evaluate the capacity of 13-93B1 glass scaffolds to repair bone defects in vivo. Scaffolds with a “trabecular” microstructure, similar to that of dry human trabecular bone, were prepared using a foam replication technique. The conversion of the glass to HA was studied in simulated body fluid (SBF) in vitro. The capacity of the scaffolds to support bone regeneration in vivo was evaluated using a rabbit femoral defect model. Since growth factors are often required to stimulate bone formation [12], the capacity of the scaffolds to serve as a carrier for platelet-rich plasma (PRP) was also evaluated in segmental defects created in the diaphysis of rabbit radii. 2. Materials and methods 2.1. Preparation of borosilicate (13-93B1) bioactive glass scaffolds Scaffolds of borosilicate (13-93B1) bioactive glass scaffolds were prepared using a polymer foam replication technique, as described previously [13]. This technique was used because of its ability to produce scaffolds with a microstructure similar to dry human trabecular bone. Briefly, glass frits with the composition 6Na2O–8K2O–8MgO–22CaO– 18B2O3–36SiO2–2P2O5 (mol%) were prepared by melting the required quantities of Na2CO3, K2CO3, MgCO3, CaCO3, H3BO3, SiO2, and NaH2PO4 · 2H2O (analytical grade; Sinopharm Chemical Reagent Co., Ltd., China) in a platinum crucible for 30 min at 1150 °C in air, and quenching the melt in cold water. Glass particles of size b 50 μm were obtained by crushing the glass frits with a hardened steel mortar and pestle and sieving through a stainless steel sieve. A slurry for the polymer foam infiltration step was prepared by dispersing the 13-93B1 particles in ethanol using ethyl-cellulose (EC; analytical grade, Sinopharm Chemical Reagent Co., Ltd., China) as a dispersant and binder. Cylindrical samples (5 mm in diameter × 15 mm) of a polyurethane foam (Shanghai No. 6 Plastic Co., Ltd., China), with open porosity of ~ 50 pores per inch, were immersed in the slurry to coat them with a layer of glass particles. After drying overnight at room temperature, the coated foams were heated in air for 1 h at 500 °C (heating rate = 1 °C/min) to burn off the foam and polymeric additives in the glass coating, and then for 2 h at 560 °C (heating rate = 5 °C/min) to sinter the glass particles into a dense network.
295
2.3. Evaluation of the bioactivity of 13-93B1 glass in vitro The bioactivity of the 13-93B1 glass was assessed in vitro from the conversion of as-fabricated scaffolds and disks (10 mm in diameter × 2 mm) in simulated body fluid (SBF) at 37 °C. One gram of glass was immersed in 100 ml SBF, as described in previous studies [5,14,15], and immersion times of up to 30 days were used. At selected times, the scaffolds and disks were removed from the SBF, washed twice with deionized water and then twice with ethanol, and dried at 90 °C. The weight loss of the scaffolds and disks was determined as ΔW = (Wo − Wt) / Wo, where Wo is the initial mass of the sample, and Wt is the mass after immersion for time t in the SBF. Four replicates were used for each time point, and the weight loss was determined as a mean ± standard deviation (SD). After the samples were removed at each time point, the SBF was cooled to room temperature and its pH was measured. The crystalline phase formed on the surface of the glass disks by conversion in SBF was analyzed using thin-film XRD (X'Pert Pro; PANalytical, Almelo, The Netherlands) using Ca Kα1 radiation (λ = 0.15405 nm; incident radiation = 45 kV) at a scanning rate of 0.03°/min in the 2θ range 10–80°C. The surface morphology and microstructure of the scaffolds after immersion in SBF were examined in a field-emission SEM (Quanta 200 FEG) using the conditions described earlier. 2.4. Evaluation of scaffolds in rabbit femoral head defect model Three male New Zealand white (NZW) rabbits weighing 2.5–3.0 kg were used in the experiments. The animals were obtained from the Shanghai Laboratory Animal Center (Shanghai, China, Certificate number SCXK: 2002-0010). Animal care and surgical procedures were in accordance with guidelines issued by the Department of Science and Technology of China in 2006 (Guidance Suggestions from the Act for Care and Use of Laboratory Animals). Under intramuscular anesthesia (3% pentobarbital sodium; 30 mg per kg mass of rabbit), lateral approaches were performed in both shaved front knees to expose the distal femoral diaphysis. One defect (5 mm diameter × 5 mm deep) was created in each of those femoral heads using a medium speed burr under constant irrigation with sterile saline. The defects were implanted with the scaffolds or left unfilled (control), and the wounds were sutured. At 4 and 8 weeks post-implantation, the femoral heads with the implants or unfilled defects were harvested and fixed in 10% neutral formalin–saline solution. The samples were sectioned in the center to expose the cross section of the implants, and decalcified in 10% formic acid. After the samples were dehydrated in a graded series of alcohol, rinsed in xylene, and embedded in paraffin, 5 μm thick sections were cut and stained with hematoxylin and eosin (H&E). The stained sections were examined in an optical microscope (AX80T, Olympus, Japan).
2.2. Characterization of 13-93B1 scaffolds 2.5. Evaluation of scaffolds in critical-sized segmental defects in rabbit radii The as-fabricated scaffolds were coated with Au/Pd and their microstructure was examined in a field-emission scanning electron microscope, SEM (Quanta 200 FEG; FEI Co., The Netherlands) at an accelerating voltage of 10 kV and a working distance of 7.5 mm. The porosity in the scaffold was measured using the Archimedes method, while the size distribution of the open pores was measured using a liquid extrusion porosimeter (LEP-1100 AX, Porous Materials Inc., NY) with water as the wetting liquid. According to the manufacturer's instruction, the distribution of pore size is defined as [−(dV / d log D)], where V is the pore volume in 1 g porous scaffold and D is the pore diameter. The compressive strength of cylindrical samples (6 mm in diameter × 12 mm) was measured using an Instron testing machine (Model 4881; Instron Co., Norwood, MA) at a crosshead speed of 0.5 mm/min. Five samples were tested, and the strength was expressed as a mean ± standard deviation.
Two groups of 13-93B1 scaffolds were evaluated in critical-sized segmental defects created in the diaphysis of rabbit radii: (1) asfabricated 13-93B1 scaffolds and (2) as-fabricated 13-93B1 scaffolds loaded with platelet-rich plasma (PRP). The unfilled defects served as the negative control group. Loading the scaffolds with PRP was performed as follows. Rabbits to be implanted with the PRP-loaded scaffolds were anesthetized by intravenously injecting 25 g/l pentobarbital sodium at a dose of 1 ml per kg body weight. Then 5 ml of blood was aspirated from the central arteries of the rabbit ears. The platelets in the blood were enriched by a two-step centrifugation process. First, the erythrocytes were removed at 8.40 N (centrifugal force at 2500 rev/min; 10 min) in a centrifuge (Heraeus Labofuge 300, Kendro Lab Products, Langenselbold, Germany) at 20 °C. Second, the leukocytes were sedimented at 13.10 N (centrifugal
296
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
3. Results 3.1. Characteristics of as-fabricated 13-93B1 scaffolds The microstructure of the as-fabricated 13-93B1 glass scaffolds (Fig. 1a) was similar to that of dry human trabecular bone. The scaffolds had a porosity of 78 ± 8% and a pore size in the range 400–650 μm, with an average of ~500 μm (Fig. 1b). The compressive strength of the as-fabricated scaffolds was 5.1 ± 1.7 MPa. 3.2. Degradation and conversion of 13-93B1 glass in vitro Immersion of the as-fabricated scaffolds in SBF resulted in the degradation and conversion of the smooth glass surface (Fig. 1a) to a porous layer composed of fine needle-like particles (Fig. 2a, b). Thin-film XRD of 13-93B1 glass disks immersed in SBF for 30 days showed diffraction peaks that corresponded to those of a reference HA (JCPDS 09-0432), confirming that the converted material was HA (Fig. 3). The broad peaks with a low intensity indicated that the HA was poorly crystallized or was composed of nanocrystalline particles. Fig. 4 shows the weight loss of the scaffolds and the pH of the SBF as a function of immersion time. The trend in the data, consisting of a rapid increase initially followed by a slower increase thereafter, is typically observed for the conversion of silicate, borosilicate, and borate bioactive glasses to HA [4,5]. After 30 days of immersion, the weight loss of the scaffolds was ~35%, while the pH of the SBF increased from an initial value of 7.4 to ~8.2. 3.3. Bone regeneration in rabbit femoral defects No post-operative complications or inflammation were observed at 4 or 8 weeks post-implantation. Fig. 5 shows the gross appearance of
Fig. 1. (a) SEM image of 13-93B1 scaffold prepared by a foam replication technique; (b) pore size distribution of the fabricated scaffold.
force at 4000 rev/min; 10 min) at 20 °C to obtain PRP with a platelet concentration equal to ~5.5 times the value for normal blood. A volume of 0.8 ml PRP was pipetted onto each scaffold and gelled by adding 0.2 ml thrombin before implantation. Eighteen NZW rabbits were used in these experiments. The rabbits were anesthetized by intravenously injecting 0.5 ml per kg body weight of ketamine (42.8 mg/ml) and xylazine (0.7 mg/ml). The rabbits were immobilized on their backs and the right hind limbs were shaved and disinfected with povidone–iodine. A longitudinal incision of 2.5–3 cm in the skin was made over the distal radius. Then the extensor retinaculum was separated to expose the bone. A segmental diaphyseal defect (1.5-cm long) was created with an oscillating saw under irrigation with sterile saline. The defects were implanted with 6 implants per group (as-fabricated scaffolds; as-fabricated scaffolds with PRP), while 6 unfilled defects served as the control group. At 4 or 8 weeks post-implantation, the rabbits were sacrificed by intravenous injection with an overdose of pentobarbital sodium. 2.6. Radiography and histology X-ray radiographs were taken of the right tibiae of the rabbits. After the removal of surrounding soft tissue and bone, the implants were photographed using a digital camera. The harvested implants were fixed in 10% formalin−saline solution and decalcified in 10% formic acid. After the samples were dehydrated in a graded series of alcohol and embedded in paraffin, longitudinal sections (5 μm thick) were cut and stained with H&E. The stained sections were examined in a transmitted light microscope (AX80T, Olympus).
Fig. 2. SEM images of the surface of 13-93B1 scaffold after immersion for 30 days in simulated body fluid (SBF): (a) lower magnification; (b) higher magnification.
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
297
Fig. 6 shows H&E stained sections of defects implanted with the 1393B1 glass scaffolds for 4 and 8 weeks. The unconverted glass was shown in white due to the decalcification, while the converted HA was stained in pink because of the adsorption of intracellular or extracellular proteins [7], indicative of partial conversion of the glass to HA. At 4 weeks, soft tissues grew into the HA formed by degradation and conversion of the glass, while new bone (denoted NB) infiltrated the macropores of the scaffold (denoted S). The amount of new bone increased at 8 weeks, and the new bone was well integrated with the HA that formed on the glass. A larger number of osteoblast-like cells were apparent in the tissues at 8 weeks than at 4 weeks.
3.4. Repair of segmental defects in rabbit radii
Fig. 3. (a) Thin-film XRD pattern of the surface of 13-93B1 glass disk after immersion for 30 days in SBF; (b) XRD pattern of a reference hydroxyapatite (JCPDS 09-0432).
Fig. 4. Weight loss of 13-93B1 scaffolds and pH of the solution as a function of immersion time of the scaffolds in SBF at 37 °C.
the defects implanted with the 13-93B1 scaffolds for 4 and 8 weeks and the unfilled defects (control) at 8 weeks. The defects implanted with the scaffolds showed better bone healing than the unfilled control group. Bone healing improved with implantation time for the defects implanted with the scaffolds (Fig. 5a, b). In comparison, the unfilled defects showed little new tissue formation, and the defects were still clearly evident. Though the defects were not created in the exact same position, they were still in the femoral head area. Bone response in this area could be viewed the same. However, the technique to create the defect in the femoral head should be improved in our future work.
After surgery, the animals implanted with the 13-93B1 scaffolds (without or with PRP) showed no signs of inflammation or infection, except for one animal implanted with a scaffold without PRP which showed symptoms of slight inflammation. Fig. 7 shows the appearance of segmental defects in the rabbit radii implanted with the two groups of scaffolds and the unfilled defects at 8 weeks. Both groups of scaffolds integrated with host bone, and callus formation apparently originated from both ends of the defect (Fig. 7a, b). In comparison, no bone formation was apparent in the unfilled defect (Fig. 7c). Radiographic images of the implants at 4 weeks showed a distinct radiolucent zone at the interface between the implant and the host bone (Fig. 8a, b), confirming the integration of the implants with host bone. For both groups of scaffolds, an osseous callus also formed around the implant at 4 weeks. At 8 weeks (Fig. 8a, b), the presence of the osseous callus was more evident in the defects implanted with the PRPloaded scaffolds. The radiographs also showed a decrease in the density of the material in the defect sites, particularly for the PRP-loaded implants, and the ends of the implants were integrated with host bone. The osseous callus grew faster and was more evident in the defects implanted with the PRP-loaded scaffolds than in the defects implanted with the scaffolds without PRP. For the unfilled defects, a small amount of osseous callus was observed at the edges of the defect and on the side of the ulna at 8 weeks, but the edges of the defect were still clearly visible (Fig. 8c). The H&E stained images of the defects showed that bone formation increased during the eight-week implantation period for both groups of scaffolds (with or without PRP) (Fig. 9), an observation that was consistent with the radiographic analysis described above. However, there were also differences in bone regeneration between the two groups of scaffolds. Four weeks post-implantation, while the amount of new bone (NB) appeared to be small, there appeared to be a greater amount of new bone formed in the defects implanted with the PRP-loaded scaffolds. The defects contained numerous osteoblast-like cells (arrows) that penetrated around and within the pores of the scaffold (S). Woven bone, as well as cancellous and lamellar bone (LB) were observed, particularly in the scaffolds loaded with PRP.
Fig. 5. Gross appearance of rabbit femoral head defects implanted with 13-93B1 scaffolds for (a) 4 weeks, and (b) 8 weeks; (c) unfilled defect at 8 weeks.
298
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
Fig. 6. H&E stained sections of femoral head defects implanted with 13-93B1 scaffolds for (a) 4 weeks, and (b) 8 weeks. S: scaffold; NB: new bone.
4. Discussion The architecture of the scaffold is known to affect tissue ingrowth, cell attachment and diffusion of nutrients into the defect site. Studies have shown that interconnected pores of size N100 μm are required to support bone ingrowth [16], but an increase in the size of the interconnected pores of the scaffold can lead to increased bone growth by allowing a greater migration of mesenchymal cells, osteoblasts, and neovascularization [17]. The borosilicate bioactive glass (13-93B1) used in the present study was fabricated into a “trabecular” microstructure with a porosity of 78 ± 8%, and a pore size of 400–650 μm, providing pore characteristics that are known to be favorable for bone ingrowth.
When immersed in SBF, the glass reacted with the solution to form a surface layer of HA (Figs. 2, 3), which is an indication of its bioactivity in vitro and its bioactive potential in vivo [14]. It was reported that the release of calcium and phosphate ions from the implants and subsequent formation of a biological HA layer could promote bone formation in vivo [18]. The adsorption of proteins from serum or the body fluid onto the high-surface-area HA layer is also reported to be important for new bone formation [19]. The degradation and conversion of the glass to HA resulted in a weight loss of the scaffolds and an increase in the pH of the SBF (Fig. 4). Assuming that all the CaO in the glass reacted to form HA with a stoichiometric composition, the theoretical weight loss of the fully converted 13-93B1 glass is ~65% [15]. After 30 days of immersion in SBF, the measured weight loss was ~ 35%, which indicates that the glass was only partially converted to HA. However, it is known that the conversion of 13-93B1 glass is faster than the slowly-degrading silicate 13-93 glass but also much slower than borate 13-93B3 glass [5,15]. Consequently, the use of 13-93B1 glass as a scaffold material could provide an optimal combination of conversion to HA and strength retention which cannot be readily achieved with 13-93 or 13-93B3 glass. The increase in the pH of the SBF during the conversion process is attributed to the dissolution of alkali ions (Na+ and K+) from the glass and the consumption of phosphate ions from the SBF. Since phosphoric acid is a stronger acid than silicic or boric acid, dissolution of Si and B from the glass, presumably as silicic acid and borate ions, respectively, coupled with the consumption of phosphate ions could also contribute to the pH increase. However, the amount of ions released in SBF, including Ca2+, Na+ and K+, was not measured in the present work. Scaffolds for bone repair should also have the requisite mechanical properties to support physiological loads and they should degrade in tune with new bone formation. As the scaffold degrades and loses strength [15], new bone formed within the pores of the scaffold should act to offset the decrease in strength. The compressive strength of the as-fabricated 13-93B1 scaffolds (5.1 ± 1.7 MPa) was within the range of strengths (2–12 MPa) reported for human trabecular bone. Consequently, the bioactive glass scaffolds used in this study could potentially be applied as a bone grafting material in the regeneration of non-loaded bone. A concern with scaffolds composed of borosilicate and, in particular, borate bioactive glasses is the toxicity of boron released during the degradation and conversion of the scaffolds to HA. While low concentrations of boron are considered to be beneficial for the formation of healthy bone [20,21], high boron concentrations can result in chronic toxicity [22,23]. Our previous work showed that when the amount of boron released is below a certain threshold concentration, boratebased bioactive glasses can support the proliferation of bone marrow stromal cells and MLO-A5 cells [24]. Bone regeneration is also largely dependent on the degradation of the scaffolds at an optimal rate and the formation of non-cytotoxic degradation products [25,26].
Fig. 7. Gross appearance of segmental defect sites in rabbit radii after implantation for 8 weeks with (a) 13-93B1 scaffolds, and (b) 13-93B1 scaffolds loaded with platelet-rich plasma; (c) unfilled defect at 8 weeks. S: scaffold; OC: osseous callus.
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
299
Fig. 8. X-ray radiographs of rabbit radius defect sites after implantation for 4 and 8 weeks with (a) 13-93B1 scaffolds, and (b) 13-93B1 scaffolds loaded with platelet-rich plasma; (c) unfilled defect at 4 and 8 weeks.
Two osseous defect models were used in this study to assess the performance of the 13-93B1 scaffolds in vivo. The rabbit model is commonly used for screening implant materials prior to testing in a larger animal model [2,27–29]. The non-critical-sized defect in a rabbit femoral head model has been used extensively in previous studies to rapidly assess bone healing, whereas a segmental defect in rabbit radii was used as a critical-sized defect model. The results showed that both groups of scaffolds were biocompatible and supported the formation of new bone in the osseous defects mainly by osteoconduction. However, scaffolds loaded with PRP appeared to show a greater capacity to support bone formation and healing of the defects when compared to the scaffolds without PRP. Several growth factors are present in PRP, such as transforming growth factor β1, insulin-like growth factor, and vascular endothelial growth factor, which are reported to be beneficial for stimulating bone formation. While there are conflicting reports on the role of PRP in enhancing bone healing [30–32], the greater ability of the PRP-loaded scaffolds in enhancing bone formation observed in this study confirmed the positive effect of PRP on bone healing. Presumably the growth factors provided the PRP-loaded scaffolds with an osteoinductive property, in addition to the inherent osteoconductive property of the porous scaffolds themselves, by releasing the growth factors to stimulate the
proliferation and differentiation of osteoblastic cells. Because PRP has a limited stability and a short life span, its effectiveness in bone healing is dependent on the use of a suitable carrier to provide a controlled release over the required time period of PRP stability. Based on the ability of the PRP-loaded scaffolds to enhance bone regeneration, the porous 13-93B1 scaffolds used in the present study apparently provided an effective carrier. In general, the results of the present study indicated that loading bioactive glass scaffolds with PRP could provide a promising approach for bone regeneration. However further experiments are needed to better control the release rate of the growth factors from the scaffold and the degradation rate of the bioactive glass scaffold. 5. Conclusion Borosilicate (13-93B1) bioactive glass scaffolds prepared by a polymer foam replication technique had pore characteristics (porosity = 78%; pore size = 400–650 μm) that are known to be favorable for supporting bone ingrowth. The scaffolds degraded and converted to hydroxyapatite when immersed in a simulated body fluid in vitro, indicating their bioactive potential. The scaffolds supported bone regeneration when implanted for 8 weeks in non-critical-sized defects in the femoral head of rabbits or in critical-sized segmental
Fig. 9. H&E stained sections of rabbit radius defects after implantation for 4 and 8 weeks with (a) 13-93B1 scaffolds, and (b) 13-93B1 scaffolds loaded with platelet-rich plasma; S: scaffold; NB: new bone; LB: lamellar bone. Arrows indicate osteocytes.
300
Y. Gu et al. / Materials Science and Engineering C 36 (2014) 294–300
defects in rabbit radii. Loading the 13-93B1 scaffolds with platelet-rich plasma enhanced bone regeneration in the segmental defects when compared to the 13-93B1 scaffolds without platelet-rich plasma. The combination of borosilicate 13-93B1 bioactive glass scaffolds and growth factors could provide a promising approach in bone repair. Acknowledgment This research was supported by the Shanghai Committee of Science and Technology (Grant No. 12JC1408500) and the National Natural Science Foundation (Grant Nos. 51072133, 51372170 & 51272274). References [1] L.L. Hench, J. Am, Ceram. Soc. 74 (1991) 1487–1510. [2] D.L. Wheeler, K.E. Stokes, H.M. Park, J.O. Hollinger, J. Biomed. Mater. Res. 35 (1997) 249–254. [3] D.L. Wheeler, K.E. Stokes, R.G. Hoellrich, D.L. Chamberland, S.W. McLoughlin, J. Biomed. Mater. Res. 41 (1998) 527–533. [4] W. Huang, D.E. Day, K. Kittiratanapiboon, M.N. Rahaman, J. Mater, Sci.: Mater. Med. 17 (2006) 583–596. [5] A. Yao, D. Wang, W. Huang, Q. Fu, M.N. Rahaman, D.E. Day, J. Am. Ceram. Soc. 90 (2007) 303–306. [6] Z. Xie, X. Liu, W. Jia, C. Zhang, W. Huang, J. Wang, J. Control. Release 139 (2009) 118–126. [7] X. Liu, Z. Xie, C. Zhang, H. Pan, M.N. Rahaman, X. Zhang, Q. Fu, W. Huang, J. Mater. Sci. Mater. Med. 21 (2009) 575–582. [8] X. Zhang, W. Jia, Y. Gu, W. Xiao, X. Liu, D. Wang, C. Zhang, W. Huang, M.N. Rahaman, D.E. Day, N. Zhou, Biomaterials 31 (2010) 5865–5874. [9] W.-T. Jia, X. Zhang, S.-H. Luo, X. Liu, W.-H. Huang, M.N. Rahaman, D.E. Day, C.-Q. Zhang, Z.-P. Xie, J.-Q. Wang, Acta Biomater. 6 (2010) 812–819. [10] Q. Fu, W. Huang, W. Jia, M.N. Rahaman, X. Liu, A.P. Tomsia, Tissue Eng. Part A 17 (2011) 3077–3084.
[11] Q. Fu, M.N. Rahaman, B.S. Bal, L.F. Bonewald, K. Kuroki, R.F. Brown, J. Biomed, Mater. Res. Part A 95A (2010) 172–179. [12] D.F. Williams, Biomaterials 30 (2009) 5897–5909. [13] X. Liu, A. Yao, D. Wang, W. Huang, H. Fu, D. Zhao, W. Lu, H. Pan, J. Inorg. Mater. 23 (2008) 322–326. [14] D. Zhao, W. Huang, M.N. Rahaman, D.E. Day, D. Wang, Acta Biomater. 5 (2009) 1265–1273. [15] Q. Fu, M.N. Rahaman, H. Fu, X. Liu, J. Biomed, Mater. Res. Part A 95A (2010) 164–171. [16] S.F. Hulbert, F.A. Young, R.S. Mathews, J.J. Klawitter, C.D. Talbert, F.H. Stelling, J. Biomed. Mater. Res. 4 (1970) 433–456. [17] V. Karageorgiou, D. Kaplan, Biomaterials 26 (2005) 5474–5491. [18] T. Hanawa, Y. Kamiura, S. Yamamoto, T. Kohgo, A. Amemiya, H. Ukai, K. Murakami, K. Asaoka, J. Biomed. Mater. Res. 36 (1997) 131–136. [19] X.D. Zhu, H.S. Fan, Y.M. Xiao, D.X. Li, H.J. Zhang, T. Luxbacher, X.D. Zhang, Acta Biomater. 5 (2009) 1311–1318. [20] F.H. Nielsen, Plant Soil 193 (1997) 199–208. [21] F.H. Nielsen, Nutr. Rev. 66 (2008) 183–191. [22] F.J. Murray, Regul. Toxicol. Pharmacol. 22 (1995) 221–230. [23] F.J. Murray, J. Trace Elem, Exp. Med. 9 (1996) 231–243. [24] H. Fu, Q. Fu, N. Zhou, W. Huang, M.N. Rahaman, D. Wang, X. Liu, Mater. Sci. Eng. C 29 (2009) 2275–2281. [25] E. Alsberg, H.J. Kong, Y. Hirano, M.K. Smith, A. Albeiruti, D.J. Mooney, J. Dent. Res. 82 (2003) 903–908. [26] H.-J. Sung, C. Meredith, C. Johnson, Z.S. Galis, Biomaterials 25 (2004) 5735–5742. [27] C.G. Trejo, D. Lozano, M. Manzano, J.C. Dodrio, A.J. Salinas, S. Dapía, E. Gómez-Barrena, M. Vallet-Regí, N. García-Honduvilla, J. Buján, P. Esbrit, Biomaterials 31 (2010) 8564–8573. [28] H. Liao, X.F. Walboomers, W.J.E.M. Habraken, Z. Zhang, Y. Li, D.W. Grijpma, A.G. Mikos, J.G.C. Wolke, J.A. Jansen, Acta Biomater. 7 (2011) 1752–1759. [29] P. Niemeyer, K. Szalay, R. Luginbühl, N.P. Südkamp, P. Kasten, Acta Biomater. 6 (2010) 900–908. [30] J. Iqbal, S.H. Pepkowitz, E. Klapper, Curr. Osteoporos. Rep. 9 (2011) 258–263. [31] A. Oryan, A. Meimandi Parizi, Z. Shafiei-Sarvestani, A.S. Bigham, Cell Tissue Banking 13 (4) (2012) 639–651. [32] R.E.C.M. Mooren, M.A.W. Merkx, E.M. Bronkhorst, J.A. Jansen, P.J.W. Stoelinga, Int. J. Oral Surg. 36 (2007) 626–631.
