Abstract
Background/Aim: Mineralized allogeneic bone substitutes are ideal biomaterials for bone regeneration in surgeries. Moreover, they are also suitable for releasing antibiotics, delivering initial concentrations many times higher than the minimal inhibitory concentration for relevant bacterial strains, without delaying bone formation. In the present study, the potential of sustained-release long-term antibiotic delivery using allografts was investigated. A broad spectrum of antibiotics, including gentamicin, vancomycin, rifampicin, and clindamycin, was incorporated into the bone blocks using a poly-L-lactic acid (PLLA) polymer coating.
Materials and Methods: An in-vivo model of implantation within the rabbit tibia plateau was used to track the sustained release of single/combined antibiotics for up to 120 days. Bony integration and tissue responses to the PLLA-coated antibiotic-delivery systems were analyzed at 4 months post implantation using histopathological analysis.
Results: The variant loaded with both vancomycin and rifampicin demonstrated the highest activity against methicillin-sensitive Staphylococcus aureus and Staphylococcus epidermidis. Histopathological analysis revealed that the tissue responses to the coated allogeneic bone substitutes were comparable in all study groups, with no observable histopathological differences. The coated bone blocks induced a strong foreign-body reaction, including high numbers of multinucleated giant cells and other immune cells but no material-associated bone growth.
Conclusion: Based on these results, future optimization can focus on selecting more efficient release of antibiotics and increasing the encapsulated concentration to sustain antibiotic release over 4 months, thereby improving the bacteriostatic effect in vivo. Furthermore, biocompatibility and osteoconductivity must be improved.
- Biocompatibility
- bone regeneration
- allograft
- drug delivery
- local antibiotic delivery
- rabbit tibia
- MIC90
- bone graft infection
- antibiotic susceptibility
Introduction
Bone injuries and defects caused by accidental and pathological events can significantly hinder an individual’s well-being and physical capabilities (1-4). Bone substitute materials play a crucial role in restoring bone function in various medical treatments such as oral or maxillofacial surgery and orthopedic surgery for joints (5-7). While autologous bone transplant remains the gold standard for bone substitution, its limited availability and the different side-effects associated with bone harvest has driven the development of both natural and synthetic alternatives. All these alternative bone substitute materials aim to mimic natural bone structure and function, offering biocompatibility, mechanical strength, and the ability to support bone regeneration and integration via osteoconduction (8). Among the various alternatives, mineralized allogenic bone substitute material is shown to be a suitable biomaterial to regenerate different types of bone defects (9).
However, a further challenge arises from infections of the defect site, which can further massively complicate the bony healing process and undermine the success of bone grafts. These surgical site infections (SSIs) develop within a defined timeframe (typically 30 days) after a surgical procedure. For procedures with a long-term implant, the time frame may extend up to 1 year. These infections affect either the surgical area itself or deeper tissues at the opening site (10, 11). Moreover, wounds after maxillofacial trauma surgery are particularly susceptible to infection due to contamination by the ubiquitous bacterial flora in the oral and nasal cavities. Additionally, the implantation of plates and screws during orthopedic and maxillofacial surgery favors bacterial growth (10, 12).
Studies examining SSIs following facial fracture repair using open reduction and internal fixation showed a variability in incidence ranging from 4% to 12% (12) and another study reported an overall prevalence of 4.2%, 4.3% and 7.3% for Europe, Asia and the USA, respectively (13). The variations stem from the differences in study populations, procedures, and definition of SSI occurrence. However, based on the available evidence, a reasonable expected rate of SSIs falls within 4% and 12%. In other words, SSIs remain a significant concern for procedures involving bone grafting or fixation of facial fractures. For mandibular fractures, the SSI rate is also reported to range from 2.3 to 12.9% (14). Similar SSI incidences of about 2.5% to 3.6% were reported for orthopedic surgeries (15, 16). A recent study reported an SSI rate of 12.4%, with increased risk for patients with craniofacial anomalies (SSI rate of 18.6%) (17). The impaction used for placing bone transplants can disrupt local circulation and reduce bone ingrowth (18, 19). In the event of an infection at the graft site, systemically administered antibiotics often cannot effectively reach the infected bone graft (20-22). Orthopedic implant surfaces are most commonly colonized by Staphylococcus epidermidis and Staphylococcus aureus (22). A significant reason for treatment failure is the formation of biofilms which make bacteria within the biofilm less susceptible to antibiotic treatment (23-25).
Efforts to minimize SSIs are ongoing. While surgical antibiotic prophylaxis has been a common preventive measure, recent reviews have cast doubt on its universal effectiveness. Broad-spectrum antibiotics were traditionally used for surgical antibiotic prophylaxis in maxillofacial surgery. However, reducing overall antibiotic use is crucial. Extensive antibiotic administration carries risks of side-effects, drug interactions, and the emergence of antibiotic-resistant bacteria, a significant global health threat (10, 12, 26, 27). Despite these risks, antibiotics remain necessary in many surgical procedures due to the potential severity of infections. However, a focus on judicious and evidence-based use is paramount (17).
Another problem in this area is that chronic biofilm infections can create conditions unsuitable for effective bone healing. Managing these effects requires careful control of targeted antibiotic delivery to optimize bone healing while controlling infections. Bone substitute materials loaded with antibiotics are commonly employed in orthopedic and oral/maxillofacial surgeries (28-30). Combining this approach with systemic antibiotic treatment offers several benefits. Firstly, it concentrates the antibiotics at the infection site, minimizing potential side-effects elsewhere in the body and reducing the risk of antibiotic resistance developing. Additionally, it can prevent infections by inhibiting the formation of biofilms, which can further prevent complications. A key component of this therapy is the use of a biomaterial or, ideally, a bone substitute material, that can deliver significant amounts of antibiotics and release them slowly over time reaching values above the minimum inhibitory concentration (MIC) for the target bacteria. Currently, materials such as bone cements are commonly used in orthopedic surgery to deliver antibiotics. However, research is ongoing to develop new biodegradable and osteoconductive materials.
Due to their outstanding osteoconductive properties, mineralized allogeneic bone substitutes not only represent an ideal biomaterial for bone regeneration in surgery, but are also suitable for releasing high amounts of antibiotics and delivering initial concentrations many times higher than the MIC90 for target bacterial strains, without delaying bone formation (6, 31). In this context, a study by Coraça-Huber and colleagues showed that fast release of the antibiotic was only measurable within the first days after implantation, which might prevent biofilm formation but not chronic infection (19). While the long-term effects of antibiotics in orthopedic and maxillofacial surgery remain under investigation, their current use is often unavoidable. There is also evidence antibiotics may influence bone healing. Therefore, a finely controlled, sustained-release approach delivering a potentially lower, continuous dosage might be a promising strategy to combat both infection and potential side-effects.
In the present study, the potential of sustained-release long-term antibiotic delivery using allografts was investigated. A broad spectrum of antibiotics, namely gentamicin, vancomycin, rifampicin and clindamycin, was incorporated into bone blocks by employment of a poly-L-lactic acid (PLLA) polymer-coating approach. A previously described in vivo implantation model, i.e., implantation within the rabbit tibia plateau, combined with already published methodologies was used to track the sustained release of antibiotics for up to 120 days (19). Finally, bony integration and the tissue responses to the PLLA-coated antibiotic delivery systems were analyzed at 4 months post implantation using established histological analysis methods (32).
Materials and Methods
Materials. Cubes of lyophilized human bone allografts (4 mm edge length) from the Cell+Tissuebank Austria GmbH (C+TBA, Krems, Austria) were used in this investigation.
Antibiotic PLLA coating and loading of bone blocks. Vancomycin hydrochloride (Lyomark, Oberhaching, Germany), rifampicin (EMD Millipore Corp, Darmstadt, Germany), clindamycin hydrochloride and gentamicin sulfate (Sigma-Aldrich, Steinheim, Germany), PLLA Resomer L210 S (Evonik, Essen, Germany), dichloromethane (99.9% P.A. grade; Honeywell, Offenbach, Germany) were used as antibiotics for the present study.
The above-mentioned bone blocks were used for antibiotic loading via a PLLA coating. Solutions of vancomycin, clindamycin, and gentamicin (250 mg/ml each) were prepared in sterile water. Each bone block was immersed in 1 ml of its respective antibiotic solution for 20 min. The blocks were then dried overnight at 37°C in a 24-well plate. PLLA was prepared in dichloromethane to a final concentration of 50 mg/ml then 0.5 ml of the PLLA-dichloromethane suspension was added to each dried bone block. For treatment with both vancomycin and rifampicin, a solution was prepared by dissolving 1,000 mg of rifampicin in 20 ml of dichloromethane, along with 1,000 mg of PLLA. Subsequently, 1,000 mg of vancomycin was directly homogenized into the PLLA-rifampicin suspension.
In-vitro antibiotic-release kinetic measurements. The detailed methodology of the kinetic measurements was previously described by Coraça-Huber and colleagues (19). In summary, the impregnated allografts were immersed in 2 ml of phosphate-buffered saline (pH 7.4; Sigma–Aldrich, Schnelldorf, Germany). At intervals of 1, 2, 3, 4, 5, 6, and 7 days, the solution (hereafter referred to as eluant) was fully removed for the measurement of release kinetics. This measurement was conducted via a standard agar diffusion assay using Bacillus subtilis in test agar (pH 8.0; Merck KGaA, Darmstadt, Germany). A total of 100 μl from each eluant or 10-fold serial dilutions (ranging from 10,000 to 0.01 mg/l) of each antibiotic were introduced into a central well with a diameter of 6 mm. The plates were incubated aerobically at 37°C for 24 h, the inhibition zone for each plate was measured in centimeters. A standard curve was generated using logarithmic regression and applied to estimate the concentration of each antibiotic in each eluant. The experiments were performed in triplicate.
In-vitro susceptibility tests. This procedure was also described in detail by Coraça-Huber et al. (19). Briefly, Müller-Hinton agar plates (Merck KGaA) were loaded with 10 μl of suspension of methicillin-sensitive Staphylococcus aureus (MSSA) (ATCC 29212; American Type Culture Collection, Manassas, VA, USA), methicillin-resistant Staphylococcus aureus isolated from a patient, Staphylococcus epidermidis (ATCC 12228; American Type Culture Collection) at 2×105 colony-forming units/ml (0.5 McFarland). A central hole with a diameter of 6 mm was made and subsequently filled with 100 μl of each antibiotic eluant from the bone blocks. Plates were incubated aerobically for 24 h at 37°C. After incubation, the zone of inhibition for each plate was measured, and concentrations were calculated using the standard curve described in the preceding section.
In-vivo animal model. The implantation procedure was conducted as previously described (19). In short, 21 New Zealand White rabbits (n=7 each group) from the Vivarium of the Faculty of Medicine, University of Niš, aged 4-5 months, and weighing 2.5-3 kg, were used. The animals were assigned to three study groups, treated with (i) clindamycin, (ii) gentamycin, and (iii) vancomycin combined with rifampicin.
During the trials, which were conducted at the Faculty of Medicine, University of Niš, Serbia, all animals were kept in conventional housing circumstances (12/12 h light/dark cycle) with unlimited access to standard food and tap water. Following approval by the local Ethics Commission of the Faculty of Medicine, University of Niš, Serbia (approval number 323-07-00278/2017-05/4 from 06.10.2017), the Veterinary Directorate of the Ministry of Agriculture, Forestry, and Water Management of the Republic of Serbia authorized animal experiments.
For the surgical implantation with allogeneic bone blocks, general anesthesia was achieved by intramuscular injection of 2-5 mg/kg xylazine and 25-50 mg/kg ketamine. A 15 mm incision was made in the tibia at the level of the medial condyle after the skin had been cleansed, shaved and sterilized. A 4×4 mm hole was drilled in the lateral condyle, and an antibiotic-coated allograft bone block was implanted. The fascia, muscle, and skin were all closed using resorbable sutures.
Each animal was routinely monitored and given ketoprofen (2 mg/kg) as an analgesic postoperatively. At 1, 3, or 120 days, animals (day 1: n=2, day 3: n=2, day 120: n=3 per group) were euthanized via barbiturate pentobarbital (400 mg/ml Euthasol®; GENERA, Kalinovica, Croatia) in compliance with the respective standards. After tibias were dissected, the allografts were removed for measurements of antibiotic concentration as described above. At day 120, two samples per group were removed for histological analysis including the complete tibia.
Determination of the residual antibiotic load in vivo. Human bone allografts were removed from the implanted site. After removal of the bone blocks, each block was placed centrally in Bacillus subtilis-inoculated (Merck KGaA) in test agar plates and the residual antibiotic load was determined as described above.
Histological and histopathological characterization of tissue responses. In accordance with a previously defined technique, this analysis was conducted using an Axio Scope.A1 light microscope (Carl Zeiss Microscopy GmbH, Oberkochen, Germany) (33-35). The goal of the qualitative histological assessment was to determine the cells that were integrating the biomaterial and to spot any negative responses, including necrosis or fibrotic encapsulation. A PC running Zen Core software connected to an Axiocam 208 Color was used to take photographs (both: Carl Zeiss Microscopy GmbH).
Statistical analysis. The calculations as well as the creation of the graphs were carried out using GraphPad Prism® (Version 8.1.3; GraphPad Software, Inc., La Jolla, CA, USA). The analysis included post hoc Tukey multiple comparisons tests and two-way analysis of variance. Statistical significance was defined as a p-value of less than 0.05.
Results
In-vivo antibiotic-release kinetics. The concentration of clindamycin was lower on day 1 and day 3 after implantation, whist remaining above the MIC90 for both S. epidermidis and MSSA (Figure 1). At day 120 post-implantation, the concentration of clindamycin was below the MIC90 for S. epidermidis, but above that for MSSA (Figure 1).
In-vivo release-kinetics for clindamycin. MSSA: Methicillin-sensitive Staphylococcus aureus; S. epidermidis: Staphylococcus epidermidis; MIC90: minimal inhibitory concentration required to inhibit 90% of isolates. Data are presented as mean±standard deviations.
The concentration of gentamycin also decreased on days 1 to 3 after implantation, falling below the MIC90 for both bacteria (Figure 2). In addition, at day 120 after implantation, the concentration of gentamycin was below the MIC90 for both bacteria (Figure 2).
In-vivo release-kinetics for gentamycin. MSSA: Methicillin-sensitive Staphylococcus aureus; S. epidermidis: Staphylococcus epidermidis; MIC90: minimal inhibitory concentration required to inhibit 90% of isolates. Data are presented as mean±standard deviations.
The concentrations of vancomycin and rifampicin decreased on day 1 and day 3 after implantation but were above the MIC90 for both bacteria (Figure 3 and Figure 4). At day 120 after implantation, the concentration of rifampicin was just at the MIC90 concentration for S. epidermidis, but not for MSSA (Figure 3). In contrast, the concentration of vancomycin was above the MIC90 for both bacteria (Figure 4).
In vivo-release kinetics of vancomycin from the combinational coating. MSSA: Methicillin-sensitive Staphylococcus aureus; S. epidermidis: Staphylococcus epidermidis; MIC90: minimal inhibitory concentration required to inhibit 90% of isolates. Data are presented as mean±standard deviations.
In vivo-release kinetics of rifampicin from the combinational coating. MSSA: Methicillin-sensitive Staphylococcus aureus; S. epidermidis: Staphylococcus epidermidis; MIC90: minimal inhibitory concentration required to inhibit 90% of isolates. Data are presented as mean±standard deviations.
Histopathological results. The histopathological analysis revealed that in all groups, a strong inflammatory response, including mainly multinucleated giant cells, macrophages and low numbers of lymphocytes, was detectable (Figure 5). The mono- and multinucleated cells were adherent to the PLLA coating but not associated with the allogeneic bone blocks. Moreover, no signs of bone tissue regeneration were found in any of the study groups.
Representative histological images at 120 days post implantation in rabbit tibia of allogeneic bone blocks impregnated with clindamycin; hematoxylin and eosin staining. (A) Overview of the implantation bed of the allogeneic bone block (dashed line). (B) - (D) Photomicrographs of the tissue reactions to the allogenic bone graft (AB) and the poly-L-lactic acid coating (asterisks) surrounded by connective tissue (CT), with associated multinucleated giant cells (black arrowheads), lymphocytes (blue arrow), and macrophages (black arrows) attached to the poly-L-lactic acid coating structures. B: Bone tissue; FT: fatty tissue; MT: muscle tissue. Original magnifications: (A) 100×, scale bar: 5 mm; (B) 10×, scale bar: 200 μm; and (C, D) 40×, scale bar: 20 μm.
Discussion
The concept of using materials for long-term antibiotic release in orthopedic and maxillofacial surgery remains under investigation to treat chronic osteomyelitis. Different strategies to achieve sufficient local antibiotic concentrations have been tested. Antibiotic-loaded polymethylmethacrylate beads were used to deliver antibiotics locally in orthopedic surgery (36, 37). However, the main disadvantages of this approach are the need to remove the beads after application and the prolonged delivery of antibiotic concentrations that are below therapeutic levels (38). Biodegradable implants loaded with antibiotics might be able to supply sufficient local concentrations without the need for later surgical removal. Many drug-delivery systems have been studied, but the majority of them have not been able to deliver antibiotics continuously at a consistent rate for the necessary amount of time (39-41).
Mineralized allogeneic bone substitutes have been shown to be a suitable material for bone regeneration in orthopedic, trauma, or dental surgeries (42). Additionally, such bone substitute material for delivering local antibiotics can provide considerable advantages over the systemic use of antibiotics while also providing an osteoconductive scaffold (43-45). In a previous in-vivo study, it was revealed that a fast release of antibiotics from this bone substitute type was measurable within the first days after implantation, which might prevent biofilm formation but not chronic infections. Thereby, the amounts of antibiotics released and delivered concentrations were many times higher than the MIC90 for relevant bacterial strains without delaying bone formation (6, 31).
In the present study, the potential of sustained-release long-term antibiotic delivery using the same allogeneic bone graft was investigated. A broad spectrum of antibiotics, including gentamicin, vancomycin, rifampicin and clindamycin, was incorporated into the bone blocks by employment of a PLLA polymer-coating approach. The results show that the concentration of antibiotics varied over time in vivo, suggesting that their clinical application would lead to different levels of bacterial resistance in clinical use. Clindamycin, vancomycin, and rifampicin showed significant release on days 1 and 3 post-implantation. The concentrations of clindamycin and vancomycin remained high on the third day, above the MIC90 for both bacteria, while only the concentration of vancomycin was above the MIC90 for both bacteria at day 120 after implantation. This indicates that the PLLA coating can basically maintain a high, localized antibiotic release during the short-term postoperative period, which is crucial for early wound healing and infection control (46). The results of the long-term release experiments indicate that combining antibiotics may be the most effective treatment against MSSA and Staphylococcus epidermidis, as previously described (46). These findings have significant implications for future modifications, including combining different antibiotics, altering surface coatings, and enhancing local infection inhibition.
Unfortunately, the histopathological analysis revealed a strong foreign-body reaction to all implanted bone blocks in all study groups, which might thus be attributable to the PLLA coating. In this context, it has been described that PLLA, which has for example been used for bone fixation for decades, can induce a foreign-body reaction that may lead to different complications (47-50).
In this context, a foreign-body response was described as being associated with the biodegradation of PLLA (47). Interestingly, the risk of a foreign-body reaction induced by a novel anti-infective coating system, which consists of PLLA mixed with adjuvants was already described (47, 51, 52). The functional principle of the coating is also based on the creation of a reservoir for the adjuvants released during the long degradation time of PLLA. It was shown that small PLLA fragments are released during application, which raised concerns about a possible foreign-body reaction. Moreover, other studies described the occurrence of a foreign-body response to PLLA implants (47, 53), as was also shown in the present study. In the case of PLLA-based bone implants, osteolysis was described, which might be a reason for the lack of material-associated bone regeneration which was observed in the present study (54, 55). The process of PLLA degradation explains why osteolysis might occur: Lactic acid accumulates as a result of PLLA degradation and is then converted to carbon dioxide and water by the body’s mitochondria-rich tissues and the liver (47). The breakdown of lactic acid takes longer in bones than in soft tissues because of the lower blood supply and fewer mitochondria in bones. In alkaline bones, osteoclastic bone resorption is initiated by local acidosis (56).
Thus, our histopathological results are in line with the literature but might also have significant implications for future modifications as different modifications might prevent the foreign-body response and osteolysis, such as altering the percentage of crystallinity, which is inversely proportional to the molecular weight in thermoplastics, or the molecular weight of the implant (53, 57). Regarding osteolytic reactions, the literature recommends modifying crystallinity or incorporating alkaline particles, such as hydroxyapatite, into PLLA to avoid this (58-60).
The long-lasting release shown in the present study indicate that future optimization should focus on selecting more efficient release of antibiotics and improving the encapsulated concentration to sustain antibiotic release over 4 months, thereby improving the bacteriostatic effect in vivo. Furthermore, biocompatibility and osteoconductivity must be improved.
Acknowledgements
The Authors gratefully acknowledge the fundings by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) for the subproject 6 within the Research Unit FOR5250 “Permanent and bioresorbable implants with tailored functionality” (No. 449916462), the Federal Ministry of Education and Research (BMBF, Germany, FKZ: 13GW0430 A, B, C) and the State Ministry of Baden-Württemberg for Economic Affairs, Labor, and Tourism (FKZ: WM3-4332.154/78/4).
Footnotes
Authors’ Contributions
Conceptualization: MB, SN, FB and XX; methodology: MB, SN, SS, MR, SN, OJ, and MB; software: MB, SN and XX; validation: MB, JZ, SS, MR, SN and XX; formal analysis: MB, SS; MR; SN, OJ and XX; investigation: MB, JZ and XX; resources: MB, SS, SN, OJ, FB and XX; data curation: MB, SS; SN, OJ, FB and XX; writing – original draft preparation: MB, JZ, SS, MR, OJ and XX; writing – review and editing: MB, SS, SN, OJ, FB and XX; visualization: MB, JZ and XX; supervision: MB, SS, SN, OJ, FB and XX; project administration: MB, SS, SN, OJ, FB and XX; funding acquisition: MB, OJ, and XX. All Authors read and agreed to the published version of the article.
Conflicts of Interest
The Authors declare no conflicts of interest.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine-learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received January 21, 2025.
- Revision received April 21, 2025.
- Accepted May 2, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).











