Bone regeneration by the combined use of tetrapod-shaped calcium phosphate granules with basic fibroblast growth factor-binding ion complex gel in canine segmental radial defects.

The effect of tetrapod-shaped alpha tricalcium phosphate granules (Tetrabones(®) [TB]) in combination with basic fibroblast growth factor (bFGF)-binding ion complex gel (f-IC gel) on bone defect repair was examined. Bilateral segmental defects 20-mm long were created in the radius of 5 dogs, stabilized with a plate and screws and implanted with 1 of the following: TB (TB group), TB and bFGF solution (TB/f group), and TB and f-IC gel (TB/f-IC group). Dogs were euthanized 4 weeks after surgery. Radiographs showed well-placed TB granules in the defects and equal osseous callus formation in all the groups. Histomorphometry revealed that the number of vessels and volume of new bone in the TB/f-IC group were significantly higher than those in the other groups. However, no significant differences in neovascularization and new bone formation were observed between the TB/f and TB groups. Furthermore, no significant difference in the lamellar bone volume or rate of mineral apposition was observed among groups. These results suggest that increased bone formation might have been because of the promotion of neovascularization by the f-IC gel. Therefore, the combinatorial method may provide a suitable scaffold for bone regeneration in large segmental long bone defects.

Bioceramic materials, such as calcium phosphate ceramics, have been extensively used for bone defects, because of their biodegradability and osteoconductivity. However, the limitation of single use of calcium phosphate ceramics may be its lack of osteoinductivity. Therefore, the combined use of calcium phosphate ceramics with various growth factors and biocompatible scaffolds has been widely investigated [13, 14].

Blood supply plays a pivotal role in fracture healing [3, 4], and enhanced vascularity has been shown to facilitate bone regeneration [20, 21, 23]. In our previous study, we demonstrated that an ion complex gel (IC gel) consisting of collagen and a citric acid derivative induced the growth of highly vascular tissues into the gel after implantation into the subfascial space in rats, and the binding of the basic fibroblast growth factor (bFGF) to the IC gel (f-IC gel) further enhanced neovascularization [22]. We also fabricated calcium phosphate alpha tricalcium phosphate granules (called Tetrabones® [TB]), which are tetrapod shaped and 1 mm in size. When packed together, the TB form intergranular pores of an appropriate size (100–400 µm), interconnecting to facilitate cellular and vascular invasions [1].

Our recent study demonstrated that the combination of TB and f-IC gel facilitated neovascularization and new bone formation in rabbits with segmental femoral defects [5]. To predict its clinical efficacy in both veterinary and human medicine, a more elaborate histomorphological study using canines with a more closely related bone metabolism to that in humans is needed.

Therefore, the present study aimed at evaluating via bone morphometric analysis, the combination effect of TB and f-IC gel on neovascularization and bone regeneration in canine segmental radial defects.

MATERIALS AND METHODS

Preparation of tetrapod-shaped alpha tricalcium phosphate granules (TB): The procedure used in the preparation of the TB was previously described [1], and the materials (Tetrabones®) were obtained from NEXT21, K.K. (Tokyo, Japan).

Preparation of bFGF-binding ion complex gel matrix (f-IC gel): The procedure used in the preparation of the f-IC gel was previously described [18, 19, 22]. Briefly, the IC gel formed spontaneously at room temperature by mixing the same volume of alkali-treated collagen modified with citric acid derivatives (CADs) and atelocollagen derived from pig tissues (Nitta Gelatin Inc., Osaka, Japan). Before gel formation, recombinant human bFGF (Kaken Pharmaceutical Co., Ltd., Tokyo, Japan) was added to the mixed solution at a final concentration of 100 ng/ml.

Animals and surgical procedures: All the animal experiments in this study were conducted in accordance with the guidelines of the Animal Care Committee of the Graduate School of Agricultural and Life Sciences at The University of Tokyo. Five male beagle dogs, aged 13–19 months and weighing 10.5–11.3 kg, were used in the experiment.

Intravenous propofol (8 mg/kg) was used to induce anesthesia, which was maintained using isoflurane (1.5–2.5%) in oxygen. Fentanyl hydrate (10–20 µg·kg−1·hr) was administered as a constant-rate infusion throughout the surgery to induce analgesia. Lactated Ringer’s solution (10 ml·kg−1·hr) was infused, and intravenous cefazolin (20 mg/kg) was administered before the skin incision was made and every 2 hr thereafter. After sterilization, the radius shaft was exposed through a craniolateral skin incision, and a critically sized (20-mm long) mid-diaphyseal segmental defect [6] was created using an oscillating saw. A defect of the same size was created in each side of the radius. A polypropylene mesh cage (PMC), 9 mm in diameter and 28 mm in height, was fabricated to hold the TB granules in the defect, as previously described (Fig. 1) [5]. Then, the PMC containing the TB with or without the bFGF or f-IC gel was inserted to the defect and fixed with an 8-hole stainless bone plate 1.8 mm thick (custom-made buttress plate, Mizuho Co., Ltd., Tokyo, Japan) and 8 screws (φ2.7-mm stainless cortical screw, Synthes Vet, Tokyo, Japan). Both ends of the PMC were tucked between the plate and the intact bone ends. Then, the muscles, fascia lata and subcutaneous tissue were sutured. Cefazolin (20 mg/kg, subcutaneously twice a day) and buprenorphine (15 µg/kg, intramuscularly 3 times a day) were administered for 3 days after surgery. A Robert-Jones bandage was used in all the dogs 1 day after surgery.

Fig. 1.

(A) Tetrabones® (TB)-containing polypropylene mesh cage (PMC). The TB was packed into the PMC. Both ends of the PMC were tucked between the plate and intact bone ends. Scale bar=500 µm.

Five dogs with 10 radial defects were included in this study and were randomly divided into the following groups: TB group (n=4), implanted with PMC and TB; TB/f group (n=3), implanted with PMC, TB (1.2 g) and bFGF solution (100 ng/ml, 1.0 ml); and TB/f-IC group (n=3), implanted with PMC, TB and f-IC gel (100 ng/ml, 1.0 ml). In the TB/f group, 100-ng/ml bFGF/phosphate-buffered saline solution was poured into the TB during surgery. The dogs in all the groups were euthanized 4 weeks after surgery.

Fluorescent bone labeling: The dogs were intravenously injected with oxytetracycline (TC; 20 mg/kg; Terramycin/LA, Pfizer Japan, Inc., Tokyo, Japan) 14 days after surgery and calcein (CL; 20 mg/kg; Sigma-Aldrich Co., Tokyo, Japan) 21 days after surgery.

Radiography and micro-computed tomography: Lateral radiographs were obtained immediately after surgery and upon euthanasia. After removing the metallic implants, with the PMC retained, micro-computed tomographic (micro-CT) images (SMX-90CT, Shimadzu Co., Kyoto, Japan) of the defects at the midsagittal sections were obtained.

Histomorphometry: The harvested radius was fixed with 70% ethanol for 5 days, stained with Villanueva bone stain for 7 days, dehydrated in ascending grades of ethanol, defatted in an acetone-methyl methacrylate monomer mixture (1:2) and embedded in methyl methacrylate (Wako Chemicals, Odawara, Japan) without decalcification. Histological sections of 40-µm thickness were cut along the midsagittal plane from the intact proximal and distal ends of the radius and defect. The specimens were examined microscopically (BX-53, Olympus Co., Tokyo, Japan) under natural and polarized light and on fluorescence microscopy. The histomorphometric measurements were made using a semiautomatic image analyzing system (System Supply, Ina, Japan). The standard bone histomorphometric nomenclature, symbols and units described in the report of the American Society for Bone and Mineral Research Histomorphometry Nomenclature Committee [16] were used. The ratios (100%) of bone distance-tissue distance (BD/TD), bone volume-tissue volume (BV/TV) and lamellar bone volume/bone volume (LBV/BV); the number of vessels/analysis area (N.Ve/mm2) and mineral apposition rate (MAR, µm/day) were calculated for each sample. The BD/TD was calculated for each sample in the midline of the defect. The LBV/BV was calculated under polarized light. Lamellar bone was defined as the regular parallel alignment of collagen as distinguished from woven bone, which was defined as a haphazard organization of collagen fibers. N.Ve/mm2 was calculated in 3 longitudinal areas (cranial, medial and caudal) of the defect. A 20-mm2 rectangle (1 × 20 mm) in each area was analyzed. Vessels were defined as luminal structures containing red blood cells. Using fluorescence microscopy, the MAR was calculated by dividing the bone thicknesses between the pairs of lines marking the administration of the fluorescent bone markers TC and CL by the time intervals between the doses (TC-CL, days 14 and 21; CL-osteoid, days 21 and 28).

Statistical analysis: Results were expressed as mean ± standard deviation. Statistical analysis was performed using the StatView Version 5.01 software (SAS Institute Inc., Cary, NC, U.S.A.). Statistical differences were analyzed using one-way analysis of variance followed by the Tukey-Kramer test for multiple comparisons. Differences were considered statistically significant at P<0.05.

RESULTS

Surgical outcome and clinical findings: Macroscopically, the defect and PMC were stabilized well using a plate and screws without TB leakage during the surgery. All the dogs tolerated the surgical procedures and had a complete, uneventful recovery.

Radiological findings: The plain radiographs (Fig. 2A–2C) and micro-CT images (Fig. 2a–2c) showed that the TB granules were placed in the defects without leakage but with slight caudal displacement of the middle portion of the TB. Although it was difficult to distinguish the newly formed bone in the intergranular pores of the TB, osseous callus formation extending from the intact bone cortex over the implant was observed 4 weeks after surgery in each group.

Fig. 2.

(A–C) Radiographs and (a–c) micro-computed tomographic images of the radial segmental bone defects in each treatment group 4 weeks after surgery. (A, a) TB group; (B, b) TB/f group; and (C, c) TB/f-IC group. The arrows indicate the osseous callus formation extending from the intact proximal bone cortex over the defect. dR: distal radius, pR: proximal radius.

Histomorphometric findings: Figure 3 shows the histological sections for each group. In the TB/f-IC group, more newly formed bone was observed, especially from the distal radius, compared with that in the TB and TB/f groups. Figure 4 shows the histological findings from the defect in each group. As shown in Fig. 4A–4C, the histological sections under natural light revealed more extensive bone formation in the TB/f-IC group than in the TB and TB/f groups. Detailed analysis of the histological sections under polarized light revealed that the newly formed bone in each group mainly consisted of woven bone (Fig. 4D–4F). Furthermore, more neovascularization was observed in the TB/f-IC group than in the TB and TB/f groups (Fig. 4G–4I). Figure 5 shows the histological sections of the fluorescence-labeled newly formed bone with TC (yellow lines) and CL (green lines) signals in each group. The mineral apposition during certain periods after implantation was analyzed by examining the incorporation of the fluorescence labels, TC and CL, into the bones. The bone thicknesses of the 2 pairs of lines, TC-CL and CL-osteoid, did not differ grossly among the groups (Fig. 5).

Fig. 3.

(A–J) Histological sections (Villanueva bone stain) at the midsagittal plane of the TB (A), TB/f (B) and TB/f-IC groups (C) 4 weeks after surgery. dR: distal radius, pR: proximal radius, PMC: polypropylene mesh cage. The arrows indicate the distance of the distal and proximal new bone tissues from the midline of the defect. Scale bar=5 mm.

Fig. 4.

Histological findings of newly formed bone (A–C: natural light, D–F: polarized light) and neovascularization (G–I: natural light) in the TB (A, D and G), TB/f (B, E and H) and TB/f-IC groups (C, F and I) 4 weeks after surgery. TB: Tetrabones, NB: new bone, WB: woven bone. The arrowheads indicate the newly formed vessels. Scale bar=200 µm.

Fig. 5.

Histological findings of fluorescently labeled newly formed bone. The yellow and green lines indicate the tetracycline and calcein signals, respectively, in the TB (A), TB/f (B) and TB/f-IC groups (C) 4 weeks after surgery: the thicknesses (i) from day 14 to 21 and (ii) from day 21 to 28. Scale bar=50 µm.

Figure 6 shows the results of the histomorphometry using the histological findings in Figs. 4 and 5. As shown in Fig. 6 (A, B and D), BD/TD, BV/TV and N.Ve/mm2 were significantly higher in the TB/f-IC group than in the other groups 4 weeks after surgery. The values of BD/TD, BV/TV and N.Ve/mm2 in the TB/f-IC group were 61.6 ± 7.0%, 17.9 ± 2.0% and 7.3 ± 0.4, respectively; however, no significant differences in these 3 parameters were found between the TB/f and TB groups. The values of BD/TD, BV/TV and N.Ve/mm2 in the TB versus TB/f group were 36.9 ± 2.6% versus 44.1 ± 2.9%, 10.1 ± 0.3% versus 9.5 ± 1.3% and 3.7 ± 0.3 versus 4.5 ± 0.8, respectively. As shown in Fig. 6 (C and E), no significant differences in LBV/BV and MAR values were observed between the groups.

Fig. 6.

Results of the histomorphometric analyses of the newly formed bone 4 weeks after surgery. (A) The distances of the proximal and distal new bone tissues per total tissue distance (BD/TD), (B) the volume of bone tissue per volume of total tissue (BV/TV), (C) the volume of lamellar bone tissue per volume of bone tissue (LBV/BV), (D) the number of vessels per square millimeter (N.Ve/mm2) and (E) the mineral apposition rate (MAR). Values are shown as mean ± standard deviation. The asterisks indicate statistically significant differences between the groups (one-way analysis of variance followed by the Tukey-Kramer test for multiple comparison; P<0.05).

DISCUSSION

In our recent study, the combination of TB and f-IC gel facilitated neovascularization and new bone formation in a rabbit segmental femoral defect model [5]. The results also demonstrated that the combination successfully facilitated neovascularization and new bone formation within 4 weeks after surgery in critically sized (20 mm) segmental radial defects in dogs.

In the present study, the number of vessels in the defects in the TB/f-IC group was significantly higher than that in the TB and TB/f groups, whereas no significant difference was found between the TB and TB/f groups. These results suggest that the f-IC gel induced angiogenic effects in the dogs. Our previous studies demonstrated that the IC gel alone markedly induced vessel growth and that its combined use with bFGF significantly enhanced the vascularization ability of the IC gel in rats [22] and rabbits [5]. Furthermore, the enhanced neovascularization in the TB/f-IC group was accompanied by a significantly greater ingrowth of newly formed bone than that in the TB and TB/f groups, suggesting that the increased formation of new bone might be because of the promotion of neovascularization by the f-IC gel.

The IC gel was developed as an extracellular matrix (ECM)-like scaffold [22]. Vessels produced in the defect were essentially within the ECM, which functioned as a natural scaffold for the vessel structure. The IC gel has a stable three-dimensional matrix structure and is degradable to develop vascular network by a variety of proteases generally released from vascular cells and related cells. Another crucial feature of the IC gel is that it serves as a reservoir of the bFGF as an angiogenic factor. Given that bFGF is positively charged and the CAD part of the IC gel is negatively charged, the bFGF can bind to the CAD with ionic bonds, fixing itself inside the IC gel. Therefore, the bFGF stimulates neovascularization and accelerates the development of a vascular network in the gel. The concentration of bFGF (100 ng/ml) in the IC gel used in this study was the minimum level required to achieve the full effect on neovascularization [22].

bFGF is known as an osteoinductive factor [2, 17] and has been demonstrated at a single local application (100–200 µg/site) to produce osteoinduction in animal bone defects or fracture models in several species including rats, rabbits, dogs and nonhuman primates [9, 11, 12, 15]. Kawaguchi et al. [10] reported that the bFGF at high concentrations acted on osteoblastic cells and stimulated not only bone formation but also bone resorption. Recent studies demonstrated that using a low dose of the bFGF (rabbits, 1.4 µg/site; dogs, 0.15 µg/site) combined with collagen minipellets as a drug delivery system successfully facilitated bone regeneration in femoral segmental defects in rabbits and guided bone regeneration (GBR) sites in dogs [7, 8]. Although the bFGF dose (100 ng/site) in this study was much lower than that used in previous studies, the f-IC gel in this study successfully facilitated bone regeneration via the binding of the bFGF to the IC gel with ionic bonds.

In addition, the f-IC gel was compatible with various granular artificial bones, in contrast to the collagen minipellets. The TB used in this study consisted of 1-mm tetrapod-shaped granules and formed intergranular pores of an appropriate size (100–400 µm), interconnecting to facilitate cellular and vascular invasions when packed together [1]. Therefore, the combination of the TB and f-IC gel may provide an ideal scaffold for bone regeneration in large bone defects that require an osteoconductive scaffold and vascular network.

In this study, we calculated the LBV/BV and MAR to precisely evaluate the osteoinductive effect of the low-dose bFGF in the IC gel. The LBV/BV value indicates the percentage of remodeling area occupied by newly formed bone and is supposed to reflect an aspect of the bone healing process in which initially formed bone, woven bone, is gradually replaced by lamellar bone through cooperative actions of osteoblasts and osteoclasts. Meanwhile, the MAR value indicates the linear rate of production of calcified bone matrix by the osteoblasts. The present study demonstrated that neither the LBV/BV nor MAR values differed significantly among the groups, suggesting that the IC gel in combination with the bFGF facilitated increased bone regeneration through neovascularization rather than by enhancing bone turnover, although the bFGF within the IC gel might have also contributed to bone regeneration. However, longer observation periods than 4 weeks are needed to get more precisely evaluation of bone turnover, because lamellar bone appears relatively late phase of the bone healing process.

In conclusion, the present study revealed that the combination of tetrapod-shaped alpha tricalcium phosphate granules and the f-IC gel facilitated bone regeneration by inducing neovascularization in a canine segmental radial defect model. This combination may be a clinically suitable scaffold for the treatment of segmental long bone defects. However, additional long-term studies are necessary to predict its clinical efficacy.