Introduction

Critical and irregular bony defects sometimes may exceed the healing capability of osteogenic tissue resulting in delayed re-ossification or non-unions [1, 2]. In such clinical challenge, additional bone grafting techniques, bone substitutes and materials are required to promote bone healing. Autologous bone has commonly been referred as the gold standard for bone grafting procedures [3].However, it has some disadvantages associated with complications regarding harvesting procedures, such as donor site morbidity, infections, chronic pains, limited graft availability, and unpredicted graft resorption rate [4]. A xenograft such as a deproteinized bovine bone (DBB) mineral can provide a scaffold and have osteoconductive properties, leading to new bone formation [5]. The bone grafting procedure most commonly used in clinical alveolar bone augmentation surgery is the usage of mixed DBB powder and autologous bone debris, in which the proportion of autologous bone debris is usually 50–60% [6, 7].The combination of autologous bone and xenograft takes advantage of each graft, reducing the amount of autologous bone as well as decreasing the healing time compared with using bone graft substitutes alone [8].

Mechanical instability and unfavorable displacement of the graft are two major concerns when grafting in the oral environment. Since particulated autologous bone chips and xenograft granules can hardly be immobilized in the bone defect, the displacement or deformation of the graft is often found in clinical practice, which may lead to poor clinical outcomes such as delayed osteogenesis, even infection [9]. Fibrin glue (FG) is a natural sealant used to create a fibrin clot for hemostasis and/or wound healing [1, 10, 11]. It has been widely used in many surgical disciplines including cranioplasty [12], spine surgery [13, 14], and maxillofacial surgery [10, 15, 16]. Beside its hemostatic capabilities, FG has been employed as scaffolds for bone and cartilage tissue engineering [17,18,19]. It has been studied in vitro studies that the addition of fibrin network to collagen sponge increased the osteoblast differentiation in a dose-dependent way, suggesting that these materials may favor bone repair [20]. Several in vivo and clinical studies have evaluated the influence of FG on bone repair, with contradictory results [12, 21].

There are various types of commercial fibrin glues in clinic, and the composition of them is different [22, 23]. In our study, a Chinese brand fibrin glue (FIBINGLURAAS®) was adopted because it contains only fibrin and thrombin with no other additives (such as aprotinin, coagulation factor, BMP, etc.), which can eliminate the interference of other factors and is more suitable for experimental research. The effect of fibrin glue on bone regeneration in the defect region repaired with autologous bone, deproteinized bovine bone, and FG is unknown. The aim of this study was to primarily testify the feasibility of using fibrin glue to stabilize graft particulates in bone defects without compromising bone regeneration. Hence, the present pilot study was designed to test the hypothesis that fibrin glue improves the bone regeneration of the calvarial defects in rats when treated with autologous bone and deproteinized bovine bone at a ratio of 1:1.

Material and methods

Preparation for DBB and fibrin glue

Deproteinized bovine bone granules (Bio-Oss®, 0.25–1.0 mm, Geistlich Biomaterials, Wolhusen, LU, Switzerland) were filtered by a filter screen with a hole diameter of 0.75 mm then separated into single packages, each of which containing about 0.02 g Bio-Oss® in the diameter of 0.25–0.75 mm, which was confirmed under an optical microscope. This preparation was carried out in a laminar flow cabinet to keep the material sterile.

Fibrin glue (FIBINGLUEAAS, RAAS, Shanghai, China), a combination of human fibrinogen and thrombin, was prepared immediately before using according to the manufacturer’s instructions. In brief, lyophilized fibrinogen (50 mg/mL) was dissolved in 2 mL sterilized water using a syringe, gently agitated the vial until the fibrinogen was completely reconstituted that the solution should be clear and have no undissolved particles. In another flask, a solution of thrombin (500 IU/mL) was combined with calcium chloride and kept at 37 °C. Fibrinogen and thrombin with calcium chloride were placed in separate syringes connected by a Y-piece that allowed them to mix in 1:1 (v/v) immediately before implantation in bone defect.

Animal experiment and surgical procedure

The research protocol was submitted to and approved by the Ethical Committee for Animal Research at the Faculty of Wenzhou Medical University, Zhejiang, China (catalogue number: wydw2017-0021).

Twenty-one male Sprague–Dawley rats weighting 250–300 g were obtained from the Wenzhou Medical University Animal Research Unit. The rats were housed in plastic cages at the temperature of 29 ± 3 °C, under natural day/night cycle. They were allowed to adapt to the new environment for 1 week prior to the study. For this animal model, a sample size of n = 3–4 animals per group per time point is commonly adopted [24, 25]. According to the results of sample power analysis, the sample size was set to be 3–4 per group of each observation point. The follow-up times of 4 weeks and 8 weeks, respectively representing the early and later period of osteogenesis, were reasonable to be the observation time for rat calvarial defect models [12]. Hence, the rats were checked for general health, weighed, numbered, and randomized into three groups:

  1. A

    Group with fibrin [fibrin ( +)] (n = 7): DBB, autologous bone, mixed in 1:1 (w/w), and fibrin glue [0.1 mL fibrinogen (50 mg/mL) mixed with 0.1 mL thrombin (500 IU/mL)]

  2. B

    Group without fibrin [fibrin ( −)] (n = 7): DBB, autologous bone, mixed in 1:1 (w/w)

  3. C

    Group control (n = 7): without bone graft

For the random allocation of the rats to three groups at two times, we first numbered all the rats and then used the formula “ = rand()” to generate a random number list. Briefly, we first numbered all the rats and put the numbers in a row in Excel. In the right adjacent row, we used the formula “ = rand()” to generate a row of random numbers between 0 and 1, and then we sorted both row in ascending order in the random number row so that the rats were also randomly aligned.

10% chloral hydrate solution (0.3 mL/100 g body weight) (Solarbio, Beijing, China) was injected for rats to achieve appropriately anesthesia before shaving the calvaria and disinfected with 5% povidone-iodine alcoholic solution (MINSHENG, Hangzhou, China). An incision was made along the sagittal suture to expose the parietal bones. An 8-mm-in-diameter skull defect was created in the middle of skull bone with an 8 mm diameter trephine by removing both external and internal cortical and preserving the dura mater. The dental drill was set at 10,000 revolutions per minute (RPM), and the trephine tip was cooled with normal saline flushes during the burring process. The bone defects were filled with different materials according to the grouping mentioned above. The cranial bone flap was served as the source of autologous bone, which was fully crushed with a tiny needle holder to bone mortars and then mixed with DBB granules of equal quality (about 0.02 g for each defect). After the suture of the periosteum with 4–0 VICRYL rapide (Ethicon, the USA), the skin was sutured with 3–0 MERSILK (Ethicon, the USA) (Fig. 1). Following skin closure, the animals received a single dose of long-acting benzathine benzylpenicillin solution (150,000 IU/100 g body weight) (NCPC, Heibei province, China). After 4 and 8 weeks, the rats were euthanized and the calvaria were harvested and processed for micro-CT scanning and histomorphometric analyses.

Fig. 1
figure 1

Operation of creating calvarial defect and implanting repair materials. A 8 mm-diameter circular bi-cortical cut was made by trephine burr in the parietal bone. B 8 mm-diameter bi-cortical parietal bone defect came out after removing the segment. C The bone defects were filled with 50% DBB and 50% autologous bone in the presence or absence of fibrin

Full size image

Micro-CT scanning and morphometric analysis

Samples that have been fixed in 10% buffered formalin for 36 h were submitted to micro-computed tomography (micro-CT) for morphometric analysis using the SCANCO MicroCT μ100 system (SCANCO Medical, Switzerland). Images were acquired at a resolution of 10 μm, 80 kV and 100 μA and reconstructed using the imaging software (SCANCO MicroCT μ100 evaluation, SCANCO Medical, Switzerland). To clearly show native bone, including autogenous bone and new bone, the gray value was set at 0–0.075. The smoothing was set at 1, ring artifact correction was set at 5, and beam hardening correction was set at 20%. We reconstructed three-dimensional images of all the hard tissues in the defect area of the rat skull. On each micro-CT images, we manually delineated the DBB according to the morphology (DBB was a polygonal particles with pores) and much higher density of DBB than newly formed bone tissue, thereby removing the DBB from the images [26, 27], and the bone tissue (including autogenous bone and new bone) was evaluated for morphometric parameters, including bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular number (Th.N), and trabecular separation (Tb.Sp).

Histomorphometric analyses

Undecalcified specimens were fixed in 4% formaldehyde solution. Afterwards, the samples were dehydrated in a series of graded ethanol solutions and embedded in methyl methacrylate. The samples were cut along their vertical axis into 10 slices of 500-μm thickness [28] at an interval of 1 mm using a precise microtome (Leica SP 1600®, Leica Microsystems B.V, Germany). Five alternate sections from each sample were mounted separately on plexiglass holders and marked with numbers. Samples’ surfaces were smoothened using polishing disks (micro-Cut Discs, Beuhler, USA) with different paper roughness from course to fine grit, 600p, 1200p, 2500p, and 4000p respectively. The slices were mounted on plexiglass holders, polished, and surface-stained with McNeal’s tetrachrome [29], basic fuchsine, and toluidine blue O [30]. Using this protocol, newly formed bone stains deep red, cell nuclei blue, collagen fibers pink, and the calcium-phosphate coating pale red. The histological sections were photographed with light microscopy (Axioskop 40, Zeiss, Germany) at a final magnification of 200 \(\times\) and microscopy (AxioCamHR, Zeiss, Germany) at a final magnification of 1.6 \(\times\) for stereological estimation. Blinded, the percentage of new bone area (Pe.NB), DBB particles (Pe.DBB), and autologous bone (Pe.AB) were measured for each group using point counting technique [31]. Average thickness of bone defect (Th.BD) was calculated by divided total points in the defect by length of defect (total points horizontally). Average thickness of bilateral basal bone (Th.BB) was measured. Percentage of new bone formation in the center of the defect (Pe.NBc) was determined by selecting one section with the widest defect. The defect was then divided into three parts and the center was marked to count the new bone (n = 3–4).

Statistical analysis

The results of micro-CT and histomorphometric analysis are presented as mean values and standard deviation (mean ± SD). Statistical analysis was performed with GraphPad Prism 8, USA). If the difference between groups was significant (one-way ANOVA), a post hoc (Student–Newman–Keuls) was performed. Statistical significance was defined as p < 0.05.

Results

Micro-CT scanning

Skull defects were filled with mineralized tissue (including the new bone, the slow resorbable DBB, and residual autologous bone) when both DBB and autologous bone were given regardless of fibrin glue at 4 and 8 weeks. In contrast, in control groups, new bone formation was only observed at the defect’s margins but not in the center of defects at 4 and 8 weeks. Fibrin ( −) group had less distinct defect boundary in comparison to fibrin ( +) group at the same time points. After 8 weeks, the amount of hard tissue was more prominent in fibrin ( −) group than in fibrin ( +) group (Fig. 2A).

Fig. 2
figure 2

Deproteinized bovine bone and autograft increased bone formation of rat calvarial bone defect in the absence of fibrin. A 3D reconstruction images from micro-computed tomography (micro-CT). B Morphometric parameters, including percentage of new bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N) and trabecular separation (Tb.Sp) were evaluated. *p < 0.05 indicates significant difference

Full size image

The bone quality was estimated using the following parameters: BV/TV (bone volume/tissue volume, %), Tb.N (trabecular number, 1/mm), Tb.Th (trabecular thickness, mm), and Tb.Sp (trabecular separation, mm). After 4 weeks, there was no difference in BV/TV among fibrin ( +) group (24.5 ± 14.4%), fibrin ( −) group (35.1 ± 13.2%), and control group (24.3 ± 13.9%). At 8 weeks, the percentage of BV/TV in fibrin ( −) group (39.1 ± 8.8%) was significantly higher than those of fibrin ( +) group (11.3 ± 10.6%) (p = 0.008) as well as control group (7.0 ± 6.5%) (p = 0.004). At each time point, no significant difference was detected between fibrin ( +) and control group. At 4 weeks, no significant difference was seen in Tb.Th among the three groups. At 8 weeks, Tb.Th significantly increased in fibrin ( −) group (0.34 ± 0.02 mm) when compared to control group (0.10 ± 0.03 mm) (p = 0.000) and fibrin ( +) group (0.19 ± 0.05 mm) (p = 0.001). After 4 weeks, Tb.N showed no significant difference among fibrin ( +) (0.50 ± 0.12 1/mm), fibrin ( −) (0.34 ± 0.15 1/mm), and control group. At 8 weeks, Tb.N significantly increased in fibrin ( −) group (1.22 ± 0.39 1/mm) than fibrin ( +) group (0.37 ± 0.20 1/mm) (p = 0.007) and control group (0.36 ± 0.11 1/mm) (p = 0.007). There was no significant difference among the three groups in Tb.Sp. A decrease tendency of Tb.Sp was found for each group at 8 weeks (Fig. 2B, Table 1).

Table 1 Micro-CT evaluation of bone regeneration induced by DBB and autologous bone without or with fibrin in calvarial bone defects of rats at 4 and 8 weeks post-operation. The bone quality was estimated using the following parameters: BV/TV (bone volume/tissue volume, %), Tb.N (trabecular number, 1/mm), Tb.Th (trabecular thickness, mm), and Tb.Sp (trabecular separation, mm). All data are presented as mean values together with the standard deviation (mean ± SD)
Full size table

Histological analysis

Although BV/TV, Tb.Th, and Tb.N increased in defect filled with DBB and autograft in absence of fibrin, it was not clear the shard tissue was the new formed bone or the autograft particles. McNeal staining has been used to stain different tissues. DBB was referred to light pink particles with sharp margin, autograft was stained in red with aligned cancellous bone lacuna inside, and new formed bone was stained in darker red wrapping bone grafts.

After 4 weeks of healing, bone repair is visible in the central portion of the defects in fibrin ( +) and fibrin ( −) group (Fig. 3A, C). Control group showed empty defect with mild new bone formation at the defect margins. At higher magnification, fibrin ( +) and fibrin ( −) group augmented defect sites frequently showed the presence of a thin bridge of newly formed bone (woven bone) which is profoundly visible at the margins extending from basal bone as well as from the autologous bone particles. This bone was continuous with the “old bone” tissue of the defect margins as well as inside the defect and contained several DBB particles. The particles appeared to be encapsulated by connective tissue, harboring large numbers of graft particles in fibrin ( +) group, while there was less connective tissue encapsulation in fibrin ( −) group (Fig. 3, 4 weeks).

Fig. 3
figure 3

Light micrographs of the bone grafting sections in the presence or absence of fibrin at time of 4 weeks and 8 weeks post-operation at low and high magnifications. Green arrow, edge of bone defect; black arrow, new bone; red arrow, DBB; white arrow, autologous bone; asterisk (*), connective tissue wrap. Magnification bars: 1 mm

Full size image
Fig. 4
figure 4

Relative total new bone area increased in fibrin ( −) group. Percentage of area of new bone area (A), DBB area (B), autograft area (C), average thickness of bone defect (D), average thickness of basal bone (E), and percentage of new bone area in the middle of bone defect (F) were analyzed. All data are presented as mean values together with the standard deviation (mean ± SD). Asterisks indicate statistically significant differences between the groups, *p < 0.05 (n = 4 for 4 weeks group, n = 3 for 8 weeks group)

Full size image

After 8 weeks, all the groups tend to have reduced fibrous scar tissue and bone remodeling from the periphery of the defects. In fibrin ( +) group, minimum new bone formation in comparison to fibrin ( −) group is observed, where ossification was more prominent, although connective tissue was also observed in the interface between basal bone and the grafted bone. At higher magnification, autologous bone particles could not be clearly distinguished anymore, which is a sign of remodeling (Fig. 3, 8 weeks).

Histomorphometric analysis

The histological morphology of new bone was estimated using the following parameters: Pe.NB (the percentage of new bone area, %), Pe.DBB (the percentage of DBB, %), Pe.AB (the percentage of autologous bone, %), Th.BD (the average thickness of bone defect, mm), Th.BB (the average thickness of bilateral basal bone, mm), and Pe.NBc (the percentage of new bone formation in the center of the defect, %) (Table 2).

Table 2 Histomorphometric analysis of bone regeneration induced by DBB and autologous bone without or with fibrin in calvarial bone defects of rats at 4 and 8 weeks post-operation. The histological morphology of new bone was estimated using the following parameters: Pe.NB (the percentage of new bone area,%), Pe.DBB (the percentage of DBB,%), Pe.AB (the percentage of autologous bone,%), Th.BD (the average thickness of bone defect, mm), Th.BB (the average thickness of bilateral basal bone, mm), and Pe.NBc (the percentage of new bone formation in the center of the defect, %). All data are presented as mean values together with the standard deviation (mean ± SD)
Full size table

There was no significant difference in percentage of new bone area (Pe.NB) after 4 weeks among fibrin ( +) group (23.75 ± 2.87%), fibrin ( −) group (29.00 ± 6.33%), and control group (23.67 ± 7.23%). After 8 weeks, Pe.NB in fibrin ( −) group (33.67 ± 11.72%) was significantly higher than that of fibrin ( +) group (12.33 ± 3.21%) (p = 0.038) and control group (9.66 ± 8.50%) (p = 0.045). There was no significant difference between fibrin ( +) and control group (p = 0.638) (Fig. 4A).

There was no significant difference in the percentage of DBB area (Pe.DBB) between fibrin ( +) group (8.50 ± 2.65%) and fibrin ( −) group (7.00 ± 2.45%) either after 4 weeks (p = 0.043) or after 8 weeks between fibrin ( +) group (13.67 ± 5.69%) and fibrin ( −) group (9.33 ± 7.51%) (p = 0.047) (Fig. 4B). There was also no significant difference in the percentage of autologous bone (Pe.AB) between fibrin ( +) group (10.50 ± 4.36%) and fibrin ( −) group 10.25 ± 5.19%) either after 4 weeks (p = 0.943) or after 8 weeks between fibrin ( +) group (3.33 ± 2.08%) and fibrin ( −) group (8.00 ± 5.00%) (p = 0.209). At both time points, no autograft was found in control groups (Fig. 4C).

Average thickness of bone defect (Th.BD) significantly increased in fibrin ( +) (1.28 ± 0.45 mm) (p = 0.043) after 4 weeks. After 8 weeks, Th.BD in fibrin ( +) group (1.45 ± 0.37 mm) (p = 0.033) and fibrin ( −) group (1.53 ± 0.05 mm) significantly increased (p = 0.000) as well. Difference of Th.BD between fibrin ( +) and fibrin ( −) was not significant at both time points (Fig. 4D).

After 4 weeks, average thickness of bilateral basal bone (Th.BB) was not significantly different in each group, whereas after 8 weeks bilateral Th.BB was significantly thicker in group fibrin ( +) (1.44 ± 0.05 mm) (p = 0.001) and group fibrin ( −) (1.57 ± 0.13 mm) (p = 0.0004) when compared with control group (Fig. 4E).

Four weeks after the grafting, new bone formation in the center of the skull defect (Pe.NBc) significantly increased in fibrin ( −) (12.05 ± 3.91%) when compared to control (0.00 ± 0.00%) (p = 0.000) and fibrin ( +) (4.79 ± 1.21%) (p = 0.005). On the contrary, there was no significant difference of Pe.NBc among the three groups after 8 weeks (Fig. 4F).

Discussion

Xenograft such as deproteinized bovine bone (DBB) and autologous bone grafts are co-administrated for grafting critical-sized bone defects that exceed the self-healing capacity. However, mechanical forces generated by oral muscle movements and chewing result in a mechanically instable microenvironment, which may lead to dispersion of bone graft particles and impairment of bone regeneration. This pilot study was designed to test the hypothesis that fibrin glue could be used as an adhesive for bone graft particles without negative effect on bone regeneration. The rat calvarial defect model (8 mm in diameter) adopted in our experiment is economical, practical, and reproducible, and can fully simulate the pathological state of cranio-maxillofacial bone defects [32, 33]. Micro-CT-based morphometric and histology-based histomorphometric analyses were carried out using rat calvarial 8 mm critical-sized defect model. At 4 weeks and 8 weeks post-surgery, fibrin ( −) group showed more bone formation when compared to fibrin ( +) group indicating that fibrin glue may inhibit osteogenesis within the first 8 weeks of bone growth.

We evaluated the effect of the clinically available fibrin glue on the efficacy of bone grafts (mixture of DBB and autologous bone) in healing the critical-size bone defects. For the rat calvarial defects, the 8-mm-in-diameter bone defects are generally accepted to be of critical size, which cannot heal spontaneously [34]. Typically, only one defect can be created per rat due to the limited area of rat calvaria. Another frequently adopted model is 5-mm-in-diameter bone defects. In comparison with the 8-mm-in-diameter bone defects, the 5-mm-in-diameter bone defect can enable the creation of two bone defects in one animal, which dramatically reduce the needed number of animals [35]. On the other hand, the 5-mm-in-diameter bone defects are of subcritical size and thus may heal without intervention [36]. Studies whose primary goal is the regeneration of bone in a defect where natural regenerative capacity no longer suffices should avoid such designs. Additionally, the potential for interactions between these two adjacent defects should be considered. Specifically for our current study, we adopted the clinically available particulate DBB and on-the-spot harvested autologous bone chips as bone-defect-healing materials and their size ranged from 0.25 to 1 mm. In our preliminary study, we adopted the 5-mm-in-diameter bone defects and found that only few materials might be placed fit into the defects, which resulted in a large heterogeneity among different defects. Therefore, we determined to adopt the 8-mm-in-diameter bone defects.

In our study, fibrin ( −) group had dominance in BV/TV, Tb.Th, and Tb.N values in comparison to fibrin ( +) and control group during the whole observation period. This suggests that fibrin glue compromises bone formation by decreasing trabecular thickness and trabecular number during the bone production phase. Furthermore, Tb.Sp showed a decrease tendency after 8 weeks of healing in all the three groups, while no significant difference of Tb.Sp among the groups was found. Our data showed that 8 weeks post-operation, the values of Tb.Th/Tb.N (0.34 mm/1.22 mm−1) in the fibrin ( −) group at this time point were significantly higher than those in the control group (0.1 mm/0.36 mm−1) and the fibrin ( +) group (0.18 mm/0.37 mm−1). In a similar pattern, Tb.Sp of fibrin ( −) group already showed a much lower mean value (0.65 mm) in comparison with those in the control group (1.36 mm) and in the fibrin ( +) group (0.97 mm), in which no statistical difference was found. Such a phenomenon might be partially explained by the less sensitivity of this parameter due to the presence of DBB in the grafts. DBB is well known to have a very low degradation rate [37]. Within the short monitoring span in this study, we assumed that DBB degraded only very mildly. Consequently, the constant presence of DBB resulted in a relatively stable separation of trabecular bone. On the other hand, the significantly higher Tb. N and Tb.Th caused certain decrease of Tb.Sp. Therefore, the joint effect of these two factors made the Tb.Sp less sensitive than Tb.N and Tb.Th, which caused the insignificantly lower level of Tb.Sp in the fibrin ( +) group than the other two groups.

Figure 2A shows the 3D reconstruction of the mineralized tissues and materials in the rat calvarial bone defects. Different from quantitative analysis, the DBB could not be excluded for 3D reconstruction; therefore, the presence of DBB made the difference of new bone not significant as the quantitative analysis. Our results indeed showed the decreases of BV/TV in the groups of control and fibrin ( +), which was not statistically significant. The decreases in both groups might be attributed to different biological mechanisms. In the control group, the skull defect was not filled with any bone-defect-filling materials, leading to a mechanically unstable microenvironment. Mild bone regeneration took place in response to acute injury in the peripheral area of the defects at 4 weeks post-operation. However, the newly formed bone might be thereafter, to some extent, resorbed by osteoclasts in this mechanically unstable microenvironment, which led to a relatively lower BV/TV at 8 weeks post-operation. As for the group of fibrin ( +), the degradation might be mediated by the activities of foreign-body giant cells [38] triggered by the presence of fibrin glue. In the fibrin ( −) group, it was also reasonable that most bone regeneration happens in the first 4 weeks since the fresh particulated autologous bone graft was the only driving force for bone regeneration. Although no significant changes in BV/TV from 4 to 8 weeks, we did observe that the trabecular numbers were greatly increased at 8 weeks, which indicated an active remodeling in this period.

The histological results showed new bone formation in grafted groups in contrast with the control group, while there was less new bone formation in fibrin ( +) group during the same time interval. Our findings showed no differences between the percentage of Bio-Oss and autograft area which prevent the discriminative evaluation of new formed bone when the amount of bone scaffold was similar regardless of fibrin (Fig. 3B, C). The compromised bone regeneration in fibrin ( +) group could be due to fibrin-induced inflammatory response [39], interference of fibrin with natural blood clot formation, foreign body reaction [40, 41], or the bone resorption ability of thrombin by boosting the osteoclastic activity [42]. As it has been observed in this study, fibrous encapsulation of grafted particles could separate the graft particles and hamper the integration of deproteinized bovine mineral particles with autologous bone and newly formed bone [43, 44]. Connective tissue which is rich in fibrinogen may enhance granulation tissue formation and deposition of collagen [45]. Fibrin glue degradation starts as early as 24 h with a half-life of 4–6 days, which means that the fibrin glue in the body will be completely absorbed within 30 days [1]. Therefore, the fibrous capsule that we observed at 8 weeks was not the residual fibrin, but probably foreign body reaction in the body. Detection of macrophages will be carried out in further study.

Fibrin immobilizes bone grafts and maintains space of osteogenesis. Average thickness of skull defect was increased in presence of fibrin since 4 weeks (Fig. 4D). This suggests fibrin glue can act as a scaffold to maintain the thickness of defect and adhere to the bone grafts. Given the high thrombin concentration in fibrin, the sealants provided short fibers and small pores to immobilize the bone graft [46]. The clustered bone grafts encapsuled by fibrin facilitated the bone regeneration as it was convenient to grasp bone graft as a whole. Increased osteogenesis space promoted osteogenic cells migration and increased new bone formation [47]. Despite the positive effect on stabilizing the bone grafts, there was a drawback of fibrin. The increased thickness of basal bone and bone defect after 8 weeks was not significantly different between fibrin ( −) and fibrin ( +) groups (Fig. 4E). Study in vitro showed that cell migration through the scaffold was increased only if lower concentration of thrombin and fibrinogen was in comparison to high concentrations of fibrinogen and/or thrombin [48, 49]. The increased thickness of basal bone and the percentage of new formed bone in the middle of bone defect were resulted from the application of bone graft rather than application of fibrin which may prevented infiltration of osteoblasts and vessel cells [46] (Fig. 4E, F). Therefore, our results demonstrate that fibrin maintains structure of bone grafts which provide osteogenic space while it is not an ideal matrix for cell invasion and osteogenesis.

Physiological and mechanical characteristics of the human clinical situation can closely be represented by animal models, but it is essential to consider the anatomic variation between human bone and each animal model when results of such studies are translated into clinical practice. Wide range of fibrinogen concentrations between 50 and 115 mg have been used in different experimental setups. In general, there is a tendency towards fibrinogen concentration between 75 and 110 mg [50]. As we know from the literatures that higher concentrations of fibrinogen could adversely affect the osteogenic potential of fibrin glue, it is obvious that more in vivo studies are needed to find the most optimal treatment regimen for bone regeneration in critical-sized defects when incorporating fibrin glue as an adhesive. Small sample size, shorter follow-up, and higher variation in the data are the limitations of this study. Further studies should be taken to see the effects of fibrin at different dosages/concentrations with larger sample size and longer healing period. Bigger animal models and a bigger sample size would be favorable. Another limitation of this study was the lack of immunohistochemical analysis of inflammatory cells (CD45, CD31, FSP1), osteoclasts (TRAP), and osteogenic markers. The main goal of this study was to primarily testify the feasibility of using clinically available fibrin glue to stabilize graft particulates (mixture of DBB and autologous bone) in bone defects without compromising bone regeneration. However, our results showed the clinically available fibrin glue might compromise the bone healing efficacy of the bone grafts. Therefore, in our subsequent study, we will try to testify other hydrogels, such as diluted fibrin glue, photocrosslinkable GelMA [51,52,53,54]. When we find the hydrogels that can both stabilize grafts and also promote their bone healing efficacy, we will further perform the immunohistochemical analysis to systematically assess the performance of the hydrogels. Furthermore, fibrin glue can be used as an injectable delivery system in bone tissue regeneration. This would make its use favorable in the clinical settings.

Conclusion

Combination of fibrin with both DBB and autograft does not additionally accelerate new bone formation. Nevertheless, fibrin is beneficial to bone autograft immobilization and adhesion. Therefore, the clinical application of fibrin to fix bone graft materials should be cautious.