Fractures in the horse industry are challenging, are frequently comminuted and can rapidly progress to fixation failure and laminitis of the contralateral limb which are common causes of death in racehorses (8). Fracture healing involves the interaction of cells, growth factors, and requires mechanical stability (2,4). To accelerate fracture healing, three major components of tissue engineering (TE) including osteogenic mesenchymal stem cells (MSC), osteoinductive bioactive molecules (growth factors), and osteoconductive biomaterials (bone grafts) are applicated alone and in combination (15).

MSC are attracting research attention for bone TE because of their ability to differentiate into bone cells (16). Platelet-rich plasma (PRP) includes α-granules that secrete growth factors such as platelet-derived growth factor, transforming growth factor-beta (TGF-β), and vascular endothelial growth factor, which enhance cell migration and regeneration (17,24). Bone morphogenetic protein 2 (BMP2) is a group of related proteins in the TGF-β superfamily that can induce osteogenesis (28,30).

Clinically desired bone grafts involve endogenous bone graft material that provides both osteoconduction and osteoinduction; however, these grafts have several disadvantages, including donor site morbidity and limited availability (28). Recently, synthetic biphasic calcium phosphate collagen composite (BCPC) grafts have been developed, and BCPC grafts can overcome those problems. Biphasic calcium phosphate (BCP), which consists of a mixture of hydroxyapatite (HA) and beta-tricalcium phosphate (β-TCP), has been widely used because its mineral composition resembles that of human bone (14,28). Moreover, when added to BCP, collagen provides a graft product that can promote bone regeneration via its ability to retain blood clots and induce cell migration (27). Therefore, BCPC is expected to obtain biological and mechanical ability that are favorable for bone formation. Research studies have also shown that in vivo implantations of BCPC have resulted in higher bone healing in both rabbit and canine in vivo models (12,13,23,31). Still, there is no evidence that BCPC promotes bone healing in large animal models.

Although TE appears to be an optimal strategy for bone regeneration, there are no reports comparing the osteogenic effects of TE approaches in horses. This study was performed to determine the bone regenerative effects of BCPC graft implantation in an equine bone defect model and to compare results with those from repeated injections of BMP2, PRP, and MSCs without bone grafts.

Animals

This randomized, blinded study included four Thoroughbred horses (male: 2, gelding: 2). They were 3.25 ± 1.26 years old and weighed 460.50 ± 60.50 kg. Before the surgical procedure, their physical conditions were assessed as normal based on physical examinations, complete blood counts, serum chemistry, and radiography. All horses were individually stalled and were monitored daily by a veterinarian. Animal experiments were approved by the Animal Ethics Committee of JeJu National University (2019).

Study design

In each horse, three cortical defects were drilled into the third metacarpal bones (MC) and metatarsal bones (MT). Two sets of six bone defects were assigned in a randomized manner to each of five treatment groups and one control group (n = 8). The five treatments groups were BCPC (n = 8), BMP2 (n = 8), MSC (n = 8), PRP (n = 8), and saline (n = 8). Injections of saline, BMP2, MSC, and PRP were made at 1, 3, and 5 weeks after bone defect surgery (Fig. 1).

Figure 1. (A) Illustration of study design. (B) Intraoperative photograph of drilled bone defect sites. (C) Photograph of taking radiographs and injecting agents under general anesthesia.

Materials

The BCPC used in this study had a 6 mm diameter and 10 mm vertical height (Osteon 3 collagen, Genoss. Co. Ltd, Suwon, Korea). It was composed of BCP (HA: β-TCP = 60:40) and porcine type I collagen (collagen volume of 47%). The BCPC had an approximate macropore size of 400-850 µm and a macroporosity of 80%.

Commercially available BMP2 (Genoss Co. Ltd., Suwon, Korea) was used in this study. A few minutes before the injection, 0.1 mg of BMP2 was diluted with 0.2 mL of distilled water.

Protocols of bone marrow aspiration, MSC culture, and cell counting used in the present study followed those previously reported (22). Reverse transcription-polymerase chain reaction (RT-PCR) and trilineage differentiation were performed as previously described to verify the multipotent effect of the MSC (22). A few minutes before injection, MSC (2 × 10⁶ cells) were suspended in 0.2 mL of saline (Isotonic Sodium Chloride Injection, Dai Han Pharm Co., Seoul, Korea).

The PRP preparation processes were similar to those in a previous protocol (11). Complete blood count and PRP were analyzed by using an automated hemocytometer (VetScan HM5; Abaxis, Inc., Union City, CA, USA). The platelet counts of the PRP used in this study were three to five times higher than that of whole blood (mean ± SD: 597 ± 270 × 10⁶ platelets/mL). A few minutes before injection, 0.2 mL of PRP were prepared in a 1 mL syringe.

Surgical procedure

Surgical procedures were similar to a previously described protocol (22). Horses were sedated intravenously with 0.05 mg/kg medetomidine hydrochloride (Equadin, Dongbang Inc., Suwonsi, Korea). A few minutes later, 0.03 mg/kg of diazepam (Diazepam Myi Amp., Myungin Pharm Co., Seoul, Korea) and 2.2 mg/kg ketamine (Yuhan Ketamine 50 Inj., Yuhan Pharm Co., Seoul, Korea) were administered intravenously. Thereafter, the trachea was intubated, and the horse was placed on a surgical table in dorsal recumbency. Subsequently, the horses were anesthetized by inhalation of isoflurane (Ifran, Hana Pharm Co., Seoul, Korea). Three skin incisions were made on each of the four limbs to expose the 3rd MC and MT. The first skin incision was 7.5-8 cm distal to the proximal end of the 3rd MC or MT. Approximately 1 cm long skin incisions were made on the palmar or plantar surface of the 3rd MC or MT at an interval of 3 cm, and bone was exposed. In each of these incisions, a bone defect was drilled (Cordless Driver 4, Stryker, Seoul, Korea) in the dorsopalmar direction, perpendicular to the bone’s long axis. The diameter of the drilled holes was 5 mm, and hole depth was 10 mm. BCPC was trimmed and implanted into one of the six drilled holes in a randomized manner. Subsequently, the skin was stapled together (skin stapler; Covidien Korea Co., Korea), and bandages were applied.

After surgery, phenylbutazone and sodium salicylate (Arthridine, Virbec South Korea, Seoul, Korea) 6.6 mg/kg were intravenously administered once a day for 3 days. For antibiotics, 0.05 mL/kg penicillin G procaine-benzathin and dihydrostreptomycin sulfate (DS long-acting PPS Inj., Daesung Microbiological Labs. Co., Ltd, Seoul, Korea) were intramuscularly administered once a day for 3 days. Dressings to the operative skin areas were applied for 9 weeks after the surgery. Postoperative general clinical and lameness checks were performed once weekly.

Injection procedure

Prior to injection, the horse was premedicated intravenously with 0.02 mg/kg medetomidine hydrochloride (Equadin, Dongbang Inc., Suwonsi, Korea). Subsequently, 0.03 mg/kg of diazepam (Diazepam Myi Amp., Myungin Pharm Co., Seoul, Korea) and 2.2 mg/kg ketamine (Yuhan Ketamine 50 Inj., Yuhan Pharm Co., Seoul, Korea) were administered intravenously. Thereafter, the trachea was intubated, the horse was placed in lateral recumbency, and subsequently anesthetized by inhalation of isoflurane (Ifran, Hana Pharm Co., Seoul, Korea) in oxygen.

Injections of saline (Isotonic Sodium Chloride Injection, Dai Han Pharm Co., Seoul, Korea), BMP2, MSC, or PRP were made at 1, 3, and 5 weeks after surgery (day 0). All treatments were prepared in 1 mL syringes immediately before the horse was anesthetized. For injection, a 23 G needle (1 mL disposable syringe, Hanjin Medical Co., Busan, Korea) was inserted into the selected bone defect with digital radiographic guidance (Galaxy R, Medien International Co., Gyeonggi-do, Korea). Subsequently, wounds were dressed and bandaged.

Radiographic evaluation

Radiography was conducted before the operation (pre), on the day of the surgery (day 0), and at 1, 5 and 9 weeks after the operation to evaluate the healing of the bone defect sites. Radiographic characteristics were 70 kVp, 6 mAs with a lateromedial view of the 3rd MC and MT using digital radiography (Galaxy R, Medien International Co., Gyeonggi-do, Korea). Radiographic data were assessed using an image reading system (Blade BMD, Medien International Co., Gyeonggi-do, Korea). Opacity within the bone defect region was evaluated by using a scoring system (0: none, 1: low permeability of bone defect site, 2: <20%, 3: < 30%, 4: <50%, 5: <70%, 6: <90% opacity) (26).

Quantitative computed tomography

Quantitative computed tomography (QCT) analysis was performed using a previously described protocol (22,26). All horses were anesthetized intravenously with 2.2 mg/kg ketamine (Yuhan Ketamine 50 Inj., Yuhan, Seoul, Korea) and euthanized intravenously with 0.1 mL/kg embutramide, mebezonium iodide, tetracaine hydrochloride, and dimethylformamide (T-61, Hansoo Pharm Co., Gyeonggi-do, Korea) at 9 weeks after the surgery. Each of the four limbs was dissected from the hock or carpal joints and examined using a Somatom Emotion16 CT system (Siemens Ltd., Seoul, Korea). Computed tomography (CT) data were processed using three-dimensional image processing software (Syngo CT, Siemens Ltd., Seoul, Korea). Obtained images were sliced into three central sagittal 0.6 mm wide pieces. The QCT value (Hounsfield Unit) of the region of interest (4.0 mm in height, 7.0 mm in width) was obtained. This represents the relative radiodensity value of the human compact bone as 1000. The mean QCT value of the assessed images was used in subsequent analyses.

Micro-computed tomography

Bone defect specimens were examined using a micro-computed tomography (µCT) system (130 kV and 60 µA; Sky-Scan 1173, SkyScan, Aartselaar, Belgium). The obtained µCT images were used to evaluate the quality of bone filling in each defect.

Histopathological and histomorphometric evaluation

Histopathological examination was conducted by applying previously described methods (22). Tissue sections were stained using hematoxylin-eosin (HE), and images were examined to evaluate histological changes in each treatment group. Histomorphometric analysis was performed to measure the new bone ratio (%) with an image-processing system (Photoshop 7.0, Adobe Systems, San Jose, CA, USA).

Statistical analysis

Statistical analysis of the data was performed using SPSS 12.0 software (SPSS, Illinois, USA). Data are presented as mean ± standard deviation values. Normality of all data was checked using the Shapiro-Wilk test. The radiography results were analyzed using the nonparametric Friedman test followed by the Wilcoxon signed-rank test. For multiple comparisons among the groups in the QCT and histomorphometric results, one-way ANOVA method test and Tukey-HSD post hoc procedure were performed. Statistical significance was considered at p < 0.05.

Radiographic evaluation

The radiographic results are presented in Table 1 and Fig. 2. Radiographic scores were significantly higher in the BCPC group than in the control and saline groups at three weeks after surgery (p < 0.05). At five weeks after surgery, radiographic scores were significantly higher in the BCPC group than in the control, saline, and PRP groups (p < 0.05), while at 7 and 9 weeks after surgery, radiographic scores in the BCPC group were significantly higher than in all of the other groups (p < 0.05).

Table 1 . Median (range) radiographic scores at different time points after surgery.

	Week
	0	1	3	5	7	9
Control	0 (0-0)	1 (0-1)	1 (1-2)	2 (2-3)	3 (2-3)	4 (3-4)
Saline	0 (0-0)	0 (0-1)	1 (1-2)	2 (1-3)	3 (2-4)	4 (3-4)
BCPC	0 (0-0)	1 (1-2)	2* (1-4)	4† (3-5)	5‡ (4-6)	6‡ (5-6)
BMP2	0 (0-0)	1 (0-2)	2 (1-2)	3 (2-3)	3 (3-4)	4 (3-5)
MSC	0 (0-0)	1 (0-1)	2 (1-2)	3 (2-3)	3 (2-4)	4 (3-5)
PRP	0 (0-0)	1 (0-1)	1 (1-2)	3 (2-3)	3 (2-4)	4 (3-5)

High scores denote good results..

*Significantly higher (p < 0.05) than the control, saline groups. †Significantly higher (p < 0.05) than the control, saline, and PRP groups. ‡Significantly higher (p < 0.05) than the control, saline, BMP2, MSC, and PRP groups..

Figure 2. Lateromedial radiographs of the surgically created bone defects in the control group, BMP2 group, BCPC group, MSC group, and PRP at 1, 5, and 9 weeks after surgery. Scale bar: 2 mm.

QCT evaluation

The QCT results are summarized in Fig. 3A. Higher values denote higher radiopaque density compared to the human compact bone taken as 1000. The results show that CT values were significantly higher in the BCPC group than in the other groups (p < 0.001); in addition, they were significantly higher in the BMP2 group than in the control and saline groups (p < 0.05).

Figure 3. Graph of quantitative computed tomography (QCT) scores (A) and results of μCT examination in the control (B), saline (C), BCPC (D), BMP2 (E), MSC (F), and PRP (G) groups at 9 weeks after surgery. *p < 0.05 and **p < 0.0001 between the indicated groups. BCPC group was significantly higher (**p < 0.0001) than the other groups. BMP2 group was significantly higher (*p < 0.05) than in the control and saline groups. Scale bar: 2 mm.

µCT evaluation

The amount of bone that filled the surgical defects was greater in the BCPC and BMP2 groups than in the control group (Fig. 3B, C, E). The µCT results that evaluated bone regeneration showed a similar trend to those of the histopathological evaluation.

Histopathological and histomorphometric evaluation

At 9 weeks after surgery, the new bone ratio (%) which is shown in Fig. 4 was significantly higher in the BMP2 group than in the control and saline groups (p < 0.05). The results of bone regeneration based on histopathological evaluation showed a similar trend to that of the µCT evaluation.

Figure 4. Graph of histomorphometric evaluation of new bone formation ratio (A) and results of histopathological examination of the defects in the control (B), saline (C), BCPC (D), BMP2 (E), MSC (F), and PRP (G) groups at 9 weeks after surgery. *p < 0.05 between the indicated groups. BMP2 group was significantly higher (*p < 0.05) than in the control and saline groups. Scale bar: 2 mm.

This study was performed to determine the bone regenerative effects of BCPC graft implantation in an equine bone defect model and to compare results with those from repeated injections of BMP2, PRP, and MSCs without bone grafts. Although TE appears to be an optimal strategy for bone regeneration, no studies have compared the osteogenic effects of various TE approaches in horses.

In this study, the BMP2 group only agents that presented significantly increased new bone ratios in the histomorphometric evaluation compared to those of the control and saline groups. We injected 0.1 mg BMP2 three times in this study and the results showed those injections significantly increased new bone formation. Our results show that repeated injection of a low concentration of BMP2 can be a successful bone regeneration factor without side effects. By contrast, saline injections did not accelerate bone regeneration. The advantage of a BMP2 application is its injectability, which makes it easy to use in equine practice. However, further studies are required to determine the most effective dose and injection schedule for BMP2 treatment of horses.

In an equine bone defect model, BMP2 was effective at an low concentration (1 µg or 3 µg) (22). Moreover, bone regeneration following BMP2 application was shown to accelerate in a dose-dependent manner (18,22). However, a high concentration of BMP2 may result in low-quality ectopic bone and excessive callus (32). BMP2 application in a human clinical trial related to dental-related bone reconstruction, fracture, and lumbar fusion, and the dosages used in that study ranged from 0.75 mg to 100 mg (6). Based on those previous studies, we decided to inject 0.3 mg BMP2 in this study. Furthermore, when only BMP2 is used, it quickly dissolves; thus, its bone regeneration effect ceases due to its high solubility (31). In another study, a single BMP2 injection produced no significant osteogenesis compared to that of the control group (5). Therefore, several BMP2 injections were applied in this study. In conclusion, injection of BMP2 may be considered an alternative to bone graft insertion in equine clinical circumstances.

In this study, the BCPC group presented significantly increased QCT and radiographic scores compared to those of the other groups at 9 weeks after surgery, implying that the BCPC grafts had not yet been degraded. These results demonstrate that a BCPC graft could be a useful bone TE component as an osteoconductive structure in horses. However, this study had limited evaluations of bone regeneration using radiography and CT because BCPC was a solid bone graft that completely filled the bone defect and the other agents used were injected in the form of 0.2 mL of a liquid.

In the histopathological findings, successful mineralization was observed in the areas where there was contact between the bone and the BCPC graft which is consistent with findings in previous studies involving dog and rabbit bone defect models in which the biocompatibility of BCPC was confirmed with no side effects after 8 weeks of application (23,31). Furthermore, recently Food and Drug Administration approved BCPC are considered highly biocompatible and biodegradable grafts that have found applications in human dental implants (10,21). Also, osteoconductivity of BCPC was demonstrated in animal study, and BCPC has been shown to have increased tensile and shear stress, making it easier to manipulate (12).

As BCPC was believed to be a suitable TE factor for osteoconduction rather than osteoinduction, comparisons of the osteoinductive effects between BCPC and other agents (BMP2, MSC, and PRP) have not been established. Although BCPC grafts are believed to improve osteoinduction through cell adhesion and proliferation in the early state due to the presence of collagen, histomorphometric results in this study showed that BCPC itself did not have a significant effect on osteoinduction, which was consistent with previous studies (13,23,31).

PRP is rich in essential growth factors, and its application has gained interest in equine regeneration medicine, including fracture healing, because of its autologous source, ease of application, and relatively low cost (3,19). In the present study, PRP treatment produced higher QCT values and showed higher histomorphometric new bone ratio than those of the control and saline groups, although the increases were not statistically significant. These results were consistent with those following PRP use in fracture repair in humans and animals (7,19). This is because PRP is not a rich source of morphogenetic proteins and without an appropriate delivery system, the effects of bone regeneration are limited due to the short half-life of the growth factors of PRP (24). Moreover, there is no agreed protocol for PRP preparation and no consistent opinion about the amount of platelets in PRP for its application to be effective (17).

MSCs are considered an attractive bone regeneration source for fracture healing, and recent studies have focused on applying MSC with scaffold material in humans and horses (1,9,22,29). However, no studies have compared the bone regeneration effect of locally repeated MSC injections in an equine bone defect model. In the present study, although there was no statistical significance, the MSC group had higher QCT values and showed higher histomorphometric new bone ratio than those of the control and saline groups. This is consistent with results in previous studies, although in those studies the MSCs were loaded in scaffold material (22). Furthermore, a recent review of MSC clinical studies indicated little individual merit for MSC use in fractures in humans (9). Another study suggested that locally implanted preosteoblasts could migrate to other sites (16). However, since the MSCs were used only in small amounts (2 × 10⁶ cells/ injection) in this study and there were no carriers for them to reside in, further studies are needed.

We believe that the injected BMP2 and PRP did not interact with neighboring defect sites because of the short biological half-lives of these factors. Moreover, histological evaluations revealed no evidence of inflammatory reaction in any of the defect sites.

Equine bone defect models have been used to evaluate various combinations of TE factors, although only a small number of studies have been reported so far (18,22,26). The present study evaluated bone regeneration of various TE factors, and the results will be useful in developing future fracture treatment strategies in horses. Most previous bone TE results were from small animal models, and the findings were not fully translatable into human and horse clinical applications (20). Moreover, large animal models can indicate practical approaches to human clinical utilization since musculoskeletal diseases of horses resemble those seen in humans (25).

This is the first report to compare the osteogenic effects of a BCPC graft with those of repeated injections of BMP2, PRP, and MSC without bone grafts in an equine bone defect model. BCPC grafts were first found to be biologically safe and showed prolonged osteoconductivity in horses. More importantly, the present study demonstrated that the repeated injections of BMP2 without a carrier can be simple and promising TE factors for treating horses with bone fractures.

This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Korea (grant number NRF-2017R1C1B1006030).

The authors have no conflicting interests.