Osteoreparative potential of nanofibrillated cellulose, xenograft, and butvar-phenolic adhesive-based composite

Abstract

BACKGROUND

The development of bone substitutes that combine mechanical stability with high osteoregenerative potential remains a challenge in orthopedics. Previous studies demonstrated the efficacy of a composite based on nanofibrillated cellulose (NFC) and bone xenograft in rabbit models. We reasoned that given their individual properties, combining these components could yield superior outcomes and identified a suitable binding agent to achieve this. We hypothesize that a uniformly mixed composite provides superior integration and stability compared with standard bone grafting methods.

AIM

To evaluate the osteoregenerative potential, remodeling patterns, and migration resistance of NFC/xenograft/Butvar-phenolic (BP)-adhesive composite in a rat femoral defect model.

METHODS

Ninety-nine rats were randomly divided into three equal groups, and a defect (2 mm diameter, 1 mm deep) was created in the mid-diaphysis of the femur. In the first group (control), intervention was performed without defect filling. In the second group (xenograft), the defect was filled with a xenograft prepared according to the Marburgbone bank system. In the third group (composite), a homogeneous mixture was prepared by pre-mixing the components, and this material was implanted into the defect cavity. Eleven animals from each group were withdrawn from the experiment on days 30, 60, and 90, and histological, histochemical, and histomorphometric examinations of the femur at the surgical site were performed to assess bone defect regeneration, remodeling, and composite migration. An infected wound model was also established. The defect cavity of 11 animals in each of the xenograft and composite groups was inoculated with Staphylococcus aureus prior to graft filling. The animals were withdrawn from the experiment on day 30, and histological examination was performed to assess inflammation.

RESULTS

The use of a composite of NFC, xenograft, and BP adhesive improved cortical bone regeneration on days 30 and 60 relative to the control group and xenograft group (P = 0.001). On day 30, the composite underwent active remodeling in both the central and peripheral zones compared with the xenograft and exhibited better fixation within the defect site compared with both control group and xenograft group, including less intramedullary and extraosseous migration. This trend persisted on days 60 and 90. In the infected wound model, composite implantation was associated with reduced severity of inflammatory injury compared with the xenograft, including osteonecrosis (P = 0.001) and polymorphonuclear granulocyte infiltration (P = 0.001).

CONCLUSION

The composite based on NFC, xenograft, and BP adhesive exhibits osteoconductive and osteoinductive potential for bone defect repair in a rat femur injury model. The results demonstrate high biocompatibility and a low incidence of complications, including composite migration. In a wound infected with Staphylococcus aureus, the composite provides favorable early resistance to infection-related injury.

Key Words: Bone regeneration; Biocomposite; Nanofibrillated cellulose; Bone graft; Marburg bone bank system

Core Tip: This preclinical study investigates a tricomponent bone substitute comprised of nanofibrillated cellulose, thermally disinfected bone xenograft (Marburg Bone Bank System), and polyvinyl butyral-phenolic (Butvar-phenolic) adhesive implanted as a uniformly premixed composite in a rat femoral diaphyseal defect model. The mixed composite demonstrated superior cortical plate regeneration on days 30 and 60, significantly reduced intramedullary migration, and more pronounced scaffold remodeling compared with xenograft alone. In a Staphylococcus aureus-inoculated wound model, the composite markedly attenuated osteonecrosis and polymorphonuclear granulocyte infiltration. These results support its dual role as an osteoconductive/osteoinductive scaffold with inherent barrier resistance to infection-related bone injury.

INTRODUCTION

The repair of bone defects caused by trauma, infection, or tumors remains one of the pressing challenges in traumatology[1-8]. The “gold standard” in orthopedics is autologous bone; however, its use is limited by the need for additional surgical trauma and the risk of complications[9]. Accordingly, there is growing interest in the development of composite materials that combine organic and inorganic components to achieve a synergistic effect[10-12]. One promising material is nanofibrillated cellulose (NFC), which exhibits high strength, elasticity, and hydrophilicity[13,14], favorable barrier properties, and wide availability. At the same time, the biological inertness of nanocellulose limits its utility for osteoregeneration[15]. Bone xenograft prepared according to the Marburg Bone Bank system demonstrates high osteoconductivity[16-19]; however, it loses its osteoinductive properties due to thermal processing. Butvar-phenolic (BP) glue exhibits strong adhesion and undergoes polymerization upon contact with biological tissues[20]. Its polyvinyl butyral component has demonstrated high biocompatibility and low cytotoxicity in experimental studies[21,22]. The formaldehyde component carries potential toxicity risks, particularly with prolonged exposure, which warrants further investigation[23-26].

Essential requirements for composite materials include reducing displacement risk[27], resistance to ischemic and necrotic injury, and antimicrobial resistance[28]. These factors underscore the relevance of developing multicomponent composites that possess high osteoconductive potential, reliable fixation within the defect site, and resistance to infection.

The aim of the present study was to perform a comparative histological and histomorphometric evaluation of bone regeneration, migration, remodeling, and resistance to infection-related injury of a composite based on NFC, a xenograft, and BP adhesive in comparison with bone xenograft in a rat femoral defect model.

MATERIALS AND METHODS

Preparation of the composite

The synthesis methodology and physicochemical characteristics of NFC with dicalcium phosphate dihydrate (CaHPO₄·2H₂O) have been previously described[13].

BP adhesive is a single-component thermosetting blend of phenol-formaldehyde resole resin and PVB. The latter is synthesized via the acetalization of polyvinyl alcohol with n-butyraldehyde in an acidic medium.

The cyclic acetal groups, alternating with vinyl alcohol residues, provide the polymer with high adhesion to various surfaces and excellent elasticity. Within the adhesive, the phenol-formaldehyde resin forms a rigid cross-linked network upon curing. The degradation profile of the BP adhesive is characterized by a low decomposition rate[29].

The BP adhesive was provided as a viscous solution, containing 1.65 g of BP adhesive per 100 g of adhesive. This type of medical adhesive previously underwent preclinical and clinical trials in accordance with the Canadian Environmental Protection Act, 1999. The results concluded that the substances within the adhesive were not harmful to human health and were not released into the environment at levels detrimental to the ecosystem[30]. The commercial name of the final product cannot be disclosed to avoid potential conflicts of interest with the manufacturer. The bone xenograft was procured from the bone tissue bank at Professor H.Z. Makazhanov Multidisciplinary Hospital in Karaganda, utilizing the certified technology of the Marburg Bone Bank System, sourced from human femoral heads.

The comparative characteristics of the individual components of the biocomposite have been described in several publications, where the efficacy of NFC and bone xenograft have been demonstrated independently[13,31,32]. The combined biocomposite was prepared under absolute sterile conditions in the vivarium’s operating room immediately before implantation. For the infected bone wound group, Staphylococcus aureus (ATCC 43300) inoculum was prepared from the third passage of a Staphylococcus aureus test culture on soy broth (trypticase soy broth) and casein digest soy agar (trypticase soy agar). The suspension density was adjusted to 5 × 10⁹ CFU/mL. A buffered saline solution with 15% gelatin added to increase viscosity was used as the diluent.

Animals and surgical procedures

A total of 132 healthy adult Wistar rats with a mean age of 5 months and a mean body weight of 350 ± 50 g were selected for the experiment. The animals were housed in the vivarium of Karaganda Medical University, with four animals of the same sex per cage, under a 12-hour light/dark cycle, controlled temperature, and ad libitum access to food and water.

The study was conducted in accordance with the European Convention for the Protection of Vertebrate Animals used for Experimental and Other Scientific Purposes (No. 123, 1986) and Directive 2010/63/EU of the European Parliament on the protection of animals. The study protocol was approved by the Local Ethics Committee of Karaganda Medical University (No. 18).

Ninety-nine rats (50 males and 49 females) were randomly divided into three groups of 33 animals each. In each animal, a defect was created in the middle third of the femoral diaphysis as previously described[13].

Under Zoletil anesthesia (0.1 mg/kg; Virbac), the hair overlying the thigh was removed, and the skin was disinfected three times with antiseptic solution. A surgical incision approximately 20 mm long was then made along the lateral aspect of the thigh to expose the femur. A round bone defect (2 mm diameter, 1 mm deep) was created in the mid-diaphysis of the femur using a PIT PMG150-C engraver with a 2 mm drill bit[33,34].

In first group, surgical intervention was performed without defect filling (control). In the second group, the defect was filled with a xenograft as previously described[35]. In the composite group, a homogeneous mixture was prepared by pre-mixing the components into a uniform mass in a sterile beaker. This homogeneous, paste-like material was then implanted into the defect cavity, filling the perforated opening to the level of the periosteum.

The soft tissues of the surgical wound were closed with absorbable Vicryl 5.0 sutures (Ethicon, Johnson & Johnson, United States). During the postoperative period, all animals received intramuscular ketonal (Sandoz, Salutas Pharma GmbH; 0.04 mL/kg) once daily for 3 days.

The rats were withdrawn from the experiment by anesthetic overdose on days 30, 60, and 90 after surgery[36-38]. Following withdrawal, a portion of the femur, including the surgical site and the surrounding soft tissues, was harvested, placed in 10% formalin, and submitted for histological examination.

Twenty-two rats (12 males and 10 females) were randomly divided into two equal groups of 11 animals each. In the first group, the defect was infected with a microbiological culture of Staphylococcus aureus and filled with bone graft. In the second group, the cavity was inoculated and subsequently filled with composite as previously described.

After defect formation, the surgical field was bacterially inoculated with a bacterial cell suspension. A laboratory culture of Staphylococcus aureus, provided by the Department of Microbiology of NPJSC “Karaganda Medical University”, was used as the infectious agent. The bacterial suspension (5 × 10⁹ CFU/mL) was applied in a volume of 2 mL directly into the defect cavity using a microbiological loop. On day 30, animals were euthanized similarly to those described above. The study design is presented in Figure 1.

Figure 1  Study design.

Histopathological examination

The femoral fragment containing the surgical site was fixed in 10% neutral buffered formalin for 24 hours. After fixation, the specimens were decalcified in Biodec R (Bio-Optica Milano S.p.A., Milan, Italy) for 24 hours. The decalcified bone was sectioned longitudinally in the sagittal plane and processed for histological examination according to a standard protocol. Five-micron sections were stained with hematoxylin and eosin (HE) for morphological and inflammatory assessment, and with Masson’s trichrome to evaluate newly formed bone and bone remodeling[39-44].

HE staining procedure

Tissue sections were immersed in Mayer’s hematoxylin for 15 minutes and rinsed in tap water for 5 minutes. Subsequently, the sections were stained with eosin for one minute.

Masson’s trichrome staining procedure

The sections were stained with Masson’s trichrome using a commercial kit [Trichrome Stain (Masson), Biovitrim, TU 9398-001-89079081-2012] in strict accordance with the manufacturer’s instructions. Microscopic examination of the slides was performed on a Zeiss AxioLab 4.0 microscope at 400 × magnification. Axiovision 7.2 for Windows software was used for the acquisition and analysis of micrographs.

Morphometric examination

HE staining was used for visualization of overall cell morphology and tissue structure. Masson’s trichrome staining was used for identification of newly formed bone tissue at different stages of bone defect repair and for assessment of bone remodeling. Two independent observers experienced in working with animal models performed the morphometric examination, blinded to the group assignment of each animal. The terminology used in the histomorphometric analysis conformed to the recommendations of the Nomenclature Committee on Histomorphometry of the American Association of Mineral and Bone Scientists[40].

Histomorphometric evaluation of cortical plate tissue

Histomorphometric evaluation of the cortical plate tissue was performed as previously described[18]. The evaluation was conducted within an area radially bounded by the defect margins and laterally by the native femoral bone, as well as the outer boundary of the bone graft and/or newly formed bone tissue. The evaluation was presented as a percentage of the total defect area. For each bone defect, three histological sections were analyzed, and their mean value was calculated. Tissues indicating a nonspecific reparative process, such as vessels or Haversian canals, were not included in the quantification and constituted the smallest percentage in the callus area[41,42]. The following parameters were determined: Cortical defect closure by newly formed bone tissue, measured by length and thickness (expressed in %); and tissue composition of both the cortical plate and the intramedullary space within the defect area (expressed in %).

Composite migration was defined as the displacement of the composite beyond the cortical plate defect margins into the intramedullary space (intramedullary migration) or beyond the bone tissue into the surrounding soft tissues (external/extraosseous migration). Intramedullary composite migration was assessed according to the following scale: “0 points” the composite is completely localized within the bone defect: No signs of displacement or scattering of the material. “1 point” local displacement: ≤ 25% of the material volume extends beyond the defect zone. “2 points” moderate migration: 25%-50% of the composite is outside the defect zone. “3 points” pronounced migration: > 50% of the material is outside the defect. Extraosseous migration was defined as the presence of the composite outside the bone tissue in the surrounding soft tissues and was assessed on a binary scale: Presence/absence. Composite remodeling was defined as the processes of remineralization and structural reorganization, leading to the gradual replacement of the composite with newly formed bone tissue. Composite remodeling was assessed both qualitatively (presence or absence) and quantitatively (as a percentage of the total composite area in the histological section) in the central and peripheral regions of the scaffold. Composite degradation was defined by pathological changes in the synthetic nanocellulose, including myxoid-like biodegradation of NFC, morphological evidence of necrosis and ischemic damage, and disruption of the material’s structural integrity with the formation of an amorphous mass. Composite degradation was assessed as a percentage of the composite area.

Semiquantitative Evaluation of Inflammatory Infiltrate

Semiquantitative histological analysis was performed on each histological section, from which ten fields of interest were scanned without any overlap and captured using high-resolution software with a 400 × objective. HE staining was used to assess the inflammatory infiltrate. Parameters of the biological response were evaluated according to the presence or absence of histological signs of osteomyelitis in the form of polymorphonuclear (PMN) leukocyte infiltration and bone necrosis[43,44] and were assessed as follows: “0 points” absent; “1 point” minimal (focal necrosis < 10%); “2 points” focal (focal necrotic fields occupying 11%-30% of defect area); “3 points” diffuse (extensive fields of necrosis occupying > 30% of the defect area).

PMN were classified as: Absent (0 points); rare, 1-5 per high-power field (400 ×) (1 point); moderate infiltration (2 points); diffuse infiltration (3 points). A reparative reactive lympho-macrophage environment in the defect zone was assessed as: “0 points” absent; “1 point” lymphocytes and macrophages surround < 10% of graft fragments; “2 points” 10%-30%; “3 points” > 30% of graft fragments.

Statistical analysis

Statistical analysis was performed using Statistica 10.0 and IBM SPSS Statistics 27.0 (IBM Corp., Armonk, NY, United States). Quantitative data are presented as mean ± SD for normally distributed variables, and as median (interquartile range) for non-normally distributed data. Categorical data are presented as absolute values and percentages of the total group size, n (%). The χ² statistical test with Yates’ continuity correction was applied to analyze differences between groups for qualitative characteristics. The normality of the distribution of quantitative data was assessed using the Shapiro-Wilk test. For normally distributed data, student’s t-tests for independent samples were used. For non-normally distributed data, group comparisons at each experimental time point were performed using the Mann–Whitney U test (for two groups) or the Kruskal-Wallis test (for > 2 groups). Where applicable, these were followed by Bonferroni correction for multiple testing in pairwise comparisons. Given the limitation of sample size, Fisher’s exact test method was adopted. P < 0.05 was considered statistically significant.

RESULTS

Histological and morphometric analysis of cortical plate defect closure

On day 30 (Figure 2 and Table 1), the cortical plate defect area in the control group showed only limited evidence of bone regeneration, characterized by small, discrete foci of osteogenesis with poorly defined bony trabeculae. In all specimens, the central region of the defect remained devoid to newly formed bone and was instead infiltrated by loose, vascularized connective tissue. On average, newly formed bone occupied 50.8% of the defect area. Bone coverage was less than 50% in five subjects (45.5%) and ranged from 51% to 80% in the remaining six subjects (54.5%).

Figure 2 Histological assessment of the percentage of bone tissue within the cortical plate on days 30, 60, and 90. ^aP < 0.0167 indicates statistically significant difference in the closure of the cortical plate by newly formed bone tissue between the control and xenograft composite groups; ^bP < 0.0167 indicates statistically significant difference in the closure of the cortical plate by newly formed bone tissue between the control and mixed composite groups.

Table 1 Histopathological evaluation of bone defect healing and composite/xenograft resorption and remodeling at 30 days, n (%)/mean ± SD.

	Control (n = 11)	Xenograft (n = 11)	Composite (n = 11)	P value
Closure of cortical plate defect (%)
Bone tissue
< 30	0 (0)	0 (0)	0 (0)	aP = 0.01; bP = 0.013; cP = 0.699
30-50	5 (45.5)	3 (27.3)	1 (9.1)
51-80	6 (54.5)	8 (72.7)	9 (81.8)
> 80	0 (0)	0 (0)	1 (9.1)
Cartilage tissue
< 30	11 (100)	11 (100)	11 (100)	-
30-50	0 (0)	0 (0)	0 (0)
51-80	0 (0)	0 (0)	0 (0)
> 80	0 (0)	0 (0)	0 (0)
Fibrous tissue
< 30	2 (18.1)	4 (36.4)	11 (100)	aP = 0.001; bP = 0.001; cP = 1.00
30-50	9 (81.9)	6 (54.5)	0 (0)
51-80	0 (0)	1 (9.1)	0 (0)
> 80	0 (0)	0 (0)	0 (0)
Fibroplastic barrier
No	11 (100)	11 (100)	5 (45.5)	aP = 1.00; bP = 0.012; cP = 0.063
Yes	0 (0)	0 (0)	6 (54.5)
Migration of composite/xenograft
Intramedullary		1.6 ± 0.8	0.3 ± 0.5	aP = 0.217; bP = 0.001; cP = 0.008
No displacement		1 (9.1)	8 (72.7)
Local displacement		3 (27.3)	3 (27.3)
Moderate migration		6 (54.5)	0 (0)
Pronounced migration		1 (9.1)	0 (0)
Extraosseous migration		0 (0)	0.3 ± 0.5
Absence		11 (100)	8 (72.7)	aP = 0.748; bP = 1.00; cP = 0.748
Presence		0 (0)	3 (27.3)

^aP indicates difference between control and xenograft.

^bP indicates difference between control and composite.

^cP indicates difference between xenograft and composite.

Histomorphometric evaluation was performed in accordance with previously published criteria[48].

The second group (xenograft) exhibited scattered regions of newly formed, heterogeneous bone tissue, predominantly consisting of thin, poorly mineralized trabeculae with occasional bridging connections. Bone formation was primarily localized to the periphery of the defect and regions in direct contact with xenograft fragments.

Histomorphometric analysis revealed that newly formed bone occupied an average of 57.3% of the total defect area, while cartilaginous and fibrous tissues accounted for 10.5% and 32.2%, respectively. The extent of bone formation varied among specimens, with 3 animals (27.3%) exhibiting less than 50% bone coverage and the remaining 8 animals (72.7%) demonstrating bone coverage ranging from 51% to 80% of the defect area.

In the composite group, histomorphometric analysis revealed that newly formed bone occupied an average of 70.3% of the cortical defect area. Histologically, the defect was partially bridged by a network of interconnected anastomosing bone trabeculae. The extent of bone formation varied among specimens, with 1 animal (9.1%) exhibiting less than 50% bone coverage, 9 animals (81.8%) showing 51%-80% coverage, and 1 animal (9.1%) demonstrating greater than 80% coverage. Cartilaginous and fibrous tissues occupied an average of 14.5% and 15.3% of the defect area, respectively.

On day 60 (Figure 2 and Table 2), histological examination of the unfilled control defects revealed trabecular structures composed predominantly of cartilaginous tissue with focal areas of ossification, primarily localized along the margins of the cortical defect. Histomorphometric analysis demonstrated an average defect coverage of 63.8%. The extent of defect filling was relatively consistent among specimens, with 1 animal (9.1%) exhibiting less than 50% coverage and the remaining 10 animals (90.9%) showing coverage ranging from 51% to 80%. Cartilaginous and bone tissues collectively accounted for 36.2% of the defect area.

Table 2 Histopathological evaluation of bone defect healing and composite/xenograft resorption and remodeling at 60 days, n (%)/mean ± SD.

	Control (n = 11)	Xenograft (n = 11)	Composite (n = 11)	P value
Closure of cortical plate defect (%)
Bone tissue
< 30	0 (0)	0 (0)	0 (0)	aP = 0.010; bP = 0.008; cP = 0.270
30-50	1 (9.1)	0 (0)	0 (0)
51-80	10 (90.9)	6 (54.5)	5 (45.5)
> 80	0 (0)	5 (45.5)	6 (54.5)
Cartilage tissue
< 30	11 (100)	11 (100)	11 (100)
30-50	0 (0)	0 (0)	0 (0)
51-80	0 (0)	0 (0)	0 (0)
> 80	0 (0)	0 (0)	0 (0)
Fibrous tissue
< 30	7 (63.4)	11 (100)	11 (100)	aP = 0.001; bP = 0.007; cP = 0.606
30-50	4 (36.6)	0 (0)	0 (0)
51-80	0 (0)	0 (0)	0 (0)
> 80	0 (0)	0 (0)	0 (0)
Fibroplastic barrier
No	11 (100)	11 (100)	5 (45.5)	aP = 1.000; bP = 0.012; cP = 1.000
Yes	0 (0)	0 (0)	6 (54.5)
Migration of composite/xenograft
Intramedullary		0.8 ± 0.9	0.4 ± 0.5	aP = 0.332; bP = 0.270; cP = 0.034
No displacement		5 (45.5)	7 (63.6)
Local displacement		3 (27.3)	4 (36.4)
Moderate migration		3 (27.3)	0 (0)
Pronounced migration		0 (0)	0 (0)
Extraosseous migration		0 (0)	0.1 ± 0.3
Absence		0 (0)	10 (90.9)	aP = 0.478; bP = 0.748; cP = 0.748
Presence		0 (0)	1 (9.1)

^aP indicates difference between control and xenograft.

^bP indicates difference between control and composite.

^cP indicates difference between xenograft and composite.

Histomorphometric evaluation was performed in accordance with previously published criteria[48].

In the xenograft group, the cortical defect was bridged by thin, newly formed bone trabeculae exhibiting focal intertrabecular connections, predominantly at their terminal ends. Adjacent to these regions, areas of densely mineralized, more mature bone tissues were observed, characterized by the presence of Haversian canals of varying diameters.

Histomorphometric analysis revealed that newly formed bone occupied an average of 64.9% of the cortical defect area, while cartilaginous and fibrous tissues accounted for 11.6% and 23.5%, respectively. The extent of bone formation was relatively consistent among specimens, with 1 animal (9.1%) exhibiting less than 50% bone coverage and the remaining 10 animals (90.9%) demonstrating bone coverage ranging from 51% to 80% of the defect area (Figure 3A and C).

Figure 3 Representative photomicrographs of the cortical plate defect area and the intramedullary space on day 60 after filling the bone defect with xenograft and composite. A: Xenograft. The defect zone is completely closed by thin newly formed bone tissue (dashed line), consisting of bone trabeculae with prominent lacunae containing osteocytes. Within the intramedullary space, bone xenograft fragments with signs of lysis are identified; B: Composite. The cortical plate defect is fully restored. The composite is undergoing active remodeling and is being replaced by bone tissue; C: Xenograft. The newly formed bone tissue is actively mineralizing. A fragment of remodeling xenograft is present in the intramedullary space; D: Composite. The cortical plate is completely restored. The composite is partially resorbed and is being replaced by mature bone trabeculae. Hematoxylin and eosin staining (A and B). Masson’s trichrome (C and D). Original magnification × 40, scale bar 500 μm. White triangle: Bone plate defect margin. Black triangle: Multinucleated giant cells. Orange five-pointed star: Polymorphonuclear granulocytes. Black five-pointed star: Lymphocytes and plasma cells.

In the composite group, histological examination revealed mature trabecular bone characterized by well-organized densely mineralized trabeculae with relatively uniform thickness and regular spatial arrangement. Histomorphometric analysis demonstrated that newly formed bone occupied an average of 82% closure of the cortical defect area. Bone coverage ranged from 51% to 80% in 5 animals (45.5%) and exceeded 80% in the remaining 6 animals (54.5%). Cartilaginous and fibrous tissues accounted for 8.8% and 9.2% of the defect area, respectively (Figure 3B and D).

By day 90 (Figure 2 and Table 3), newly formed bone occupied an average of 90.4% of the cortical defect area in the unfilled control group. Histological examination revealed persistent areas of incomplete bone remodeling in 1 specimen (9.1%), whereas the remaining 10 specimens (90.9%) exhibited mature, uniformly mineralized bone tissue forming a continuous bridge across the cortical defect.

Table 3 Histopathological evaluation of bone defect healing and composite/xenograft resorption and remodeling at 90 days, n (%)/mean ± SD.

	Control (n = 11)	Xenograft (n = 11)	Composite (n = 11)	P value
Closure of cortical defect (%)
Bone tissue
< 30	0 (0)	0 (0)	0 (0)	aP = 0.217; bP = 0.365; cP = 0.699
30-50	0 (0)	0 (0)	0 (0)
51-80	1 (9.1)	0 (0)	2 (18.2)
> 80	10 (91.9)	11 (100)	9 (81.8)
Cartilage tissue
< 30	11 (100)	11 (100)	11 (100)
30-50	0 (0)	0 (0)	0 (0)
51-80	0 (0)	0 (0)	0 (0)
> 80	0 (0)	0 (0)	0 (0)
Fibrous tissue
< 30	11 (100)	11 (100)	11 (100)
30-50	0 (0)	0 (0)	0 (0)
51-80	0 (0)	0 (0)	0 (0)
> 80	0 (0)	0 (0)	0 (0)
Fibroplastic barrier
No	11 (100)	11 (100)	11 (100)
Yes	0 (0)	0 (0)	0 (0)
Migration of composite/xenograft
Intramedullary		0 (0)	0.3 ± 0.5	aP = 0.010; bP = 0.300; cP = 0.065
No displacement		11 (100)	8 (72.7)
Local displacement		0 (0)	3 (27.3)
Moderate migration		0 (0)	0 (0)
Pronounced migration		0 (0)	0 (0)
Extraosseous migration			0
Absence		11 (100)	11 (100)	aP = 0.748; bP = 1.00; cP = 0.748
Presence		0 (0)	0 (0)

^aP indicates difference between control and xenograft.

^bP indicates difference between control and composite.

^cP indicates difference between xenograft and composite.

Histomorphometric evaluation was performed in accordance with previously published criteria[48].

Within the xenograft group at day 90 (Figure 2 and Table 3), non-osseous tissue accounted for 8.5% of the defect area, comprised of 3.5% cartilaginous and 5.0% fibrous tissue. The remaining 91.5% was closed by bone tissue. Histological evaluation identified persistent zones of incomplete remodeling in a single specimen (9.1%), whereas the remaining 10 animals (90.1%) displayed mature, uniformly mineralized bone fully bridging the defect.

The composite group demonstrated complete bone restoration. Defect closure by bone tissue averaged 91.9%, with cartilaginous and fibrous tissue occupying 5.5% and 2.6% of the area, respectively.

Histological evaluation of xenograft and composite migration

On day 30 (Table 1), intramedullary migration in the control group (xenograft) was identified in 10 cases (90.9%). Among these, local migration (< 25%) was observed in 3 cases (30.0%), moderate migration (25%-50%) in 6 cases (60.0%), and extensive migration in 1 case (10.0%). No extraosseous migration of the xenograft was observed.

In the composite group, local intramedullary migration was observed in only 3 cases (27.3%), while moderate and extensive migration were absent. Extraosseous migration of the composite occurred in 3 cases (27.3%).

At day 60 (Table 2), the composite group exhibited local intramedullary migration in 4 cases (36.4%), with no instances of moderate or extensive migration recorded. Extraosseous migration of the composite material occurred in only 1 case (9.1%). By contrast, the xenograft group showed intramedullary migration in 6 cases (54.5%), evenly split between local migration (< 25%, 3 cases, 50%) and moderate migration (3 cases, 50%). No extraosseous migration was detected in this group. On day 90 (Table 3), no intraosseous or extraosseous migration was observed in either group.

Histological evaluation of resistance to bacterial inflammatory damage

On day 30 (Table 4), necrotic foci were identified in all 11 (100%) animals in the xenograft group. Among these, diffuse injury with the formation of extensive necrotic zones exceeding 30% of the defect area was observed in 2 cases (18.2%), moderate necrotic changes (10%-30%) were noted in 5 cases (45.5%), and minimal necrotic changes (less than 10%) in 4 cases (36.4%).

Table 4 Histological evaluation of resistance to bacterial inflammatory damage, mean ± SD.

	Xenograft (n = 11)	Composite (n = 11)	P value
Osteonecrosis	1.8 ± 0.8	0.4 ± 0.5	P = 0.001
Polymorphonuclear leukocyte infiltration	2.5 ± 0.5	0.9 ± 0.8	P = 0.001
Eosinophils	0	0
Lymphocyte and macrophage infiltration	1.4 ± 0.7	1.0 ± 0.6	P = 0.243
Bacterial contamination	0.2 ± 0.4	0

P indicates difference between xenograft and mixed composite. Histomorphometric evaluation was performed in accordance with previously published criteria[48].

PMN infiltration was detected in all 11 (100%) cases, characterized as diffuse in 5 cases (45.5%) and moderate in 6 cases (54.5%) (Figure 4A, C and E). Persistence of lymphocytes and macrophages was observed, primarily localized perivascularly and within the transition zone between granulation tissue and mature connective tissue. In one case (9.1%), clusters of bacterial colonies were identified.

Figure 4 Micrographs of the cortical plate defect zone and the intramedullary space on day 30 after Staphylococcus aureus inoculation. A: Xenograft group. The cortical defect margin is characterized by areas of bone necrosis and pronounced inflammatory cellular infiltration; B: Сomposite group. Partial extraosseous migration of the composite beyond the defect into the surrounding soft tissues is noted. Newly formed bone tissue (yellow dashed line) is identified in the central zone of the cortical defect, represented by bone trabeculae completely covering the defect. A zone of fibrous tissue with cellular infiltration is detected at the periphery of the defect (∆); C: Xenograft group. Osteocyte lacunae are predominantly empty, and bone trabeculae are fragmented. Diffuse infiltration by polymorphonuclear granulocytes and marked destruction are observed around the necrotic masses. A distinct demarcation zone of inflammatory cells is visible between the necrotic area and the preserved bone tissue; D: Сomposite group. A zone of fibrous tissue with vascularization, pronounced macrophage-histiocytic infiltrate, and single polymorphonuclear granulocytes are observed at the periphery of the composite; E: Xenograft group. Diffuse bone necrosis, massive clusters of polymorphonuclear granulocytes, foci of tissue breakdown, and nuclear detritus; F: Сomposite group. Fragments of pale-stained, homogeneous masses corresponding to nanofibrillated cellulose composite are identified undergoing lysis by macrophages and multinucleated foreign body giant cells. Hematoxylin and eosin staining. Original magnification × 40, scale bar 500 μm (A and B); original magnification × 100, scale bar 200 μm (C and D); original magnification × 200, scale bar 100 μm (E and F), White triangle: Bone plate defect margin. Black triangle: Multinucleated giant cells. Orange five-pointed star: Polymorphonuclear granulocytes. Black five-pointed star: Lymphocytes and plasma cells. Black arrow: Nanofibrillated cellulose. Orange triangle: Necrotic tissue.

In the composite group, necrotic changes were noted in 4 cases (36.4%), involving less than 10% of the defect area. PMN infiltration was identified in 7 cases (63.6%). Of these, 3 cases (42.9%) were moderate and 4 cases (57.1%) were focal. Lymphocyte and macrophage infiltration was characterized as focal and of moderate intensity (Figure 4B, D and F). Bacterial colonies were absent.

DISCUSSION

In this study, we evaluated bone tissue regeneration processes during the healing of a rat femoral diaphyseal defect using a composite material based on NFC, bone xenograft, and polyvinyl BP resin. Defects filled with bone xenograft alone and unfilled group (empty controls) served as comparison groups to evaluate bone regeneration in the presence or absence of the composite material. The efficacy of this approach was previously demonstrated by our group in a rabbit femoral defect model, providing the rationale for its inclusion as a comparator in the present study[45].

The primary objective of this study was to determine whether the complementary properties of the biocomposite components could enhance bone defect healing and improve resistance to infection. Bone regeneration was evaluated using histological and histomorphometric analyses, including quantification of newly formed bone, assessment of bone remodeling, and evaluation of composite and xenograft migration within the defect site.

The results demonstrated that the composite material outperforms the isolated xenograft in its ability to stimulate reparative osteogenesis in the cortical defect zone. On days 30 and 60, histological examination fo the experimental groups revealed active bone regeneration, characterized by the formation of increasingly organized bone trabeculae, progressive mineralization of the newly deposited bone matrix, and partial bridging of the cortical defect by newly formed bone tissue In contrast, regeneration in the control group was more heterogeneous, characterized by the formation of predominantly thin, poorly mineralized bone trabeculae with focal bridging contacts primarily localized at the periphery of the defect and in areas of contact with xenograft particles, and was accompanied by the persistence of a significant proportion of cartilaginous and fibrous tissue. These findings are consistent with previous studies where the addition of NFC to composite formulations contributed to earlier cortical plate restoration[46-48].

It is worth noting that NFC biodegradation had no adverse effect on the healing process. Both histological and histomorphometric findings pointed to ongoing active osteogenesis, with the composite being progressively replaced by newly formed bone tissue without any signs of heightened inflammatory response or tissue necrosis.

Material migration warrants separate discussion. The applied biocomposite was shown to have more stable fixation in bone tissue compared with the control group (P < 0.005). The absence of moderate or severe intramedullary migration of the mixed composite at all observational time points suggests that the structural homogeneity of the material and the high cohesion of its components ensure more reliable primary fixation and resistance to displacement in the diaphyseal defect model. The pre-mixed composite likely behaves as a more cohesive, homogeneous, and easily moldable mass, providing more secure primary stability.

Isolated application of the xenograft was accompanied by the most pronounced intramedullary migration during the early healing stages, likely reflecting its limited mechanical stability in the absence of a binding matrix. As the bone regenerate formed and remodeling processes progressed, this phenomenon was gradually neutralized, and xenograft migration was no longer detected by the late stages of observation.

One clinically notable finding concerned differences in fibroblastic barrier formation within the bone defect zone. A pronounced fibroblastic barrier at the cortical plate defect site was evident in over half of the animals in the mixed composite group on both days 30 and 60, whereas the isolated xenograft group showed no comparable barrier formation (P < 0.01). This higher incidence of fibrous barrier formation in the mixed composite group may stem from micromigration of nanocellulose fragments and their direct contact with adjacent soft tissue and periosteum. Such interactions could trigger a localized fibroblastic response, driving collagen synthesis and ultimately forming a connective tissue barrier that impedes osteogenic cell infiltration into the regeneration zone, consistent with previous findings[47]. In turn, this barrier may secondarily restrict angiogenesis within the central defect area. The resulting hypoxic environment tends to favor fibroblast activation and extracellular matrix production, while simultaneously suppressing osteoblastic activity and mineralization rates.

The absence of a fibroblastic reaction when using the isolated xenograft is probably due to the lack of contact between fine synthetic fragments and the periosteum, indicating differences in the mechanisms of tissue response to xenogeneic vs composite materials.

Interestingly, under conditions of bacterial contamination, the composite also demonstrated advantages over the control group. In the infected wound model without defect filling, pronounced purulent-necrotic inflammation was observed, with necrotic zones occupying more than 30% of the area. Conversely, in the groups using the composite, no pathological accumulations of bacteria were detected, and inflammatory changes were focal and of moderate or minimal severity. This effect may be attributed to several factors; first, filling the defect creates a mechanical barrier that prevents bacterial colonization while accelerating its vascularization and closure with granulation and bone tissue, thereby contributing to the restoration of the structural barrier. Additionally, nanocellulose itself possesses high hydrophilicity and is capable of sorbing exudate, which may facilitate wound drainage and reduce the local microbial load. These observations confirm the potential efficacy of the biomaterial not only as an osteoconductive scaffold but also as a component with barrier and potentially antiseptic functions, which is also characteristic of nanotechnology-based biocomposites[46].

Notably, the reduction in the inflammatory reaction severity in the infected model was not accompanied by signs of tissue damage or necrosis. This suggests that the observed effect is a consequence of the barrier and stabilizing properties of the composite rather than a manifestation of cytotoxic suppression of inflammation.

An advantage of this study is the use of a standardized femoral defect model in rats and a comprehensive histological and morphometric evaluation at early and late stages of regeneration.

A limitation is the exclusive use of histological methods. Furthermore, the safety evaluation of the BP adhesive used in this study is limited by the observation periods and histological criteria, necessitating further research to assess potential long-term toxic effects. These limitations should be considered when interpreting the results and planning future studies aimed at filling existing gaps in knowledge.

This study demonstrated that the combined composite based on NFC, bone xenograft, and BP adhesive possesses osteoinductive and osteoconductive potential when implanted into rat femoral defects. The bone xenograft performed an osteoconductive function, providing remineralization, angiogenesis, and guided bone tissue growth, which facilitated synthetic matrix remodeling from both the periphery and the depth of the defect. NFC fulfilled an osteoinductive function, creating a favorable microenvironment for osteogenesis during its gradual biodegradation. The BP adhesive exhibited adhesive properties, contributing to the stabilization and homogeneity of the mass.

Notably, the composite also proved advantageous over the control group under conditions of bacterial contamination. In the infected wound model without defect filling, rats developed marked purulent-necrotic inflammation, with necrotic areas covering more than 30% of the site. In contrast, animals treated with the composite showed no pathological bacterial accumulation, and inflammatory changes remained focal, ranging from minimal to moderate severity.

Several factors may account for this effect. First, filling the defect establishes a mechanical barrier against bacterial colonization while promoting faster vascularization and closure via granulation and bone tissue, thereby restoring structural integrity. Second, nanocellulose’s inherent hydrophilicity allows it to absorb exudate, which may enhance wound drainage and lower the local microbial burden. Together, these findings support the biomaterial’s potential not only as an osteoconductive scaffold but also as a barrier component with possible antiseptic properties.

Importantly, the reduced severity of inflammation in the infected model occurred without any evidence of tissue damage or necrosis, suggesting that this effect reflects the composite’s barrier and stabilizing properties rather than cytotoxic suppression of the inflammatory response.

One limitation of this study is its reliance on histological methods. Additionally, safety assessment of the BP adhesive was constrained by the observation periods and histological criteria used, indicating a need for further research into potential long-term toxicity. These constraints should be taken into account both when interpreting the current findings and when designing future studies to address remaining knowledge gaps.

Overall, this study demonstrated that the composite, comprised of NFC, bone xenograft, and BP adhesive, exhibits osteoinductive and osteoconductive potential when implanted into rat femoral defects. Within this composite, the bone xenograft served an osteoconductive role, supporting remineralization, angiogenesis, and guided bone tissue growth, which in turn facilitated remodeling of the synthetic matrix from both the periphery and the depth of the defect.

Another limitation of this study is the lack of microbiological assessment of the infected wound, as well as the absence of clinical evaluation of the animals in the postoperative period. Although expanding the in vivo evaluation criteria of the composite would provide a more objective analysis, this study yields substantial new data regarding the specific application features and biological interaction of the composite material.

CONCLUSION

The present study demonstrated that a composite based on NFC, a xenograft, and BP adhesive possesses osteoconductive and osteoinductive potential for the regeneration of femoral bone defects in a rat model. The results indicate that the material is biocompatible with a low incidence of complications including migration. In addition, it undergoes active remodeling, which is more pronounced when the mixed composite is used. Moreover, when bacterial wound contamination was present, these composites displayed marked barrier and antimicrobial effects, effectively inhibiting bacterial colonization and preventing the onset of purulent-necrotic inflammation.