Localized mandibular infection affects remote in vivo bioreactor bone generation

Abstract

Mandibular reconstruction requires functional and aesthetic repair and is further complicated by contamination from oral and skin flora. Antibiotic-releasing porous space maintainers have been developed for the local release of vancomycin and to promote soft tissue attachment. In this study, mandibular defects in six sheep were inoculated with 10⁶ colony forming units of Staphylococcus aureus; three sheep were implanted with unloaded porous space maintainers and three sheep were implanted with vancomycin-loaded space maintainers within the defect site. During the same surgery, 3D-printed in vivo bioreactors containing autograft or xenograft were implanted adjacent to rib periosteum. After 9 weeks, animals were euthanized, and tissues were analyzed. Antibiotic-loaded space maintainers were able to prevent dehiscence of soft tissue overlying the space maintainer, reduce local inflammatory cells, eliminate the persistence of pathogens, and prevent the increase in mandibular size compared to unloaded space maintainers in this sheep model. Animals with an untreated mandibular infection formed bony tissues with greater density and maturity within the distal bioreactors. Additionally, tissues grown in autograft-filled bioreactors had higher compressive moduli and higher maximum screw pull-out forces than xenograft-filled bioreactors. In summary, we demonstrated that antibiotic-releasing space maintainers are an innovative approach to preserve a robust soft tissue pocket while clearing infection, and that local infections can increase local and remote bone growth.

Keywords: In vivo bioreactors; Large animal model; Local antibiotic release; Mandibular repair; Osteomyelitis; Tissue engineering.

Figure 1.

Surgical photos of the mandible (A-C) or the bioreactor implantation (D-F). The diastema was exposed from the most anterior molar to the exit of the mental nerve (A), a defect was created from the superior border to the nerve canal (B), after inoculation a space maintainer was plated into the defect site (C). The periosteum was carefully elevated from the rib before to removal of the rib to preserve the periosteum (D), the bioreactor was filled with morselized rib (AG, left) or Bio-Oss® (XG, right) (E), prior to implantation and suturing into place (F).

Figure 2.

Space maintainer characterization and data collection at postoperative week 0 (surgery), 1, 2, 4, and 9 (euthanasia). In vitro characterization of vancomycin-loaded space maintainer over 9 weeks (n=3) (A). Normalized white blood cell count for untreated (blank space maintainer) or treated (vancomycin-loaded space maintainer) animals was determined (B). Palpable enlargement and large dehiscence were observable in the untreated mandible (C,left) but not the treated mandible (D,left) or the contralateral controls (C,right, D,right). The incision site (E), the oral mucosa adjacent to the space maintainer (F), and the oral flora (G) were monitored throughout the study. * indicates significant difference, p < 0.05. Data represent mean ± standard deviation.

Figure 3.

MicroCT analyses of mandibles. Untreated mandibles (A) show areas of osteolysis as well as new bone formation. Treated mandibles (B) show maintenance of space with the space maintainer located within the initial defect site. Cross-sections of the mandibles at the lines indicated in A show the remodeling that has occurred anterior (C), within (D), and posterior (E) to the defect site. These cross-sections refer to the locations of the histological slides. Screws are indicated by *. Scale bar = 15 mm (A,B) or 5 mm (C,D,E).

Figure 4.

Low magnification mandibular histology. Methylene blue-basic fuchsin staining was performed on sections from the contralateral control (A-C), untreated (D-F), and treated (G-I) mandibles. Sections were made anteriorly (A,D,G), within the defect (B,E,H), and posteriorly (C,F,I). Mandibular height (J), Width (K), and Area (L) were measured using image analysis software and normalized to the contralateral side of the same sheep. Data represent mean ± standard deviation. Those that do not share the same letter are significantly different (p < 0.05). Scale bar = 5 mm.

Figure 5.

High magnification mandibular histology. Untreated mandible shows poor formation of a fibrous capsule adjacent to the space maintainer (A) and breakdown of the oral mucosa, exposing dead bone to the oral cavity (B). Treated mandibles show bone adjacent to the space maintainer with very thin organized fibrous capsule formation and bone within pores (C) and a healthy oral mucosa over bone. High magnification of an inflammatory infiltrate within one of the specimens showing rounded cells (E). Graphs of bone-substrate score (F), pore-substrate score (G), mucosal intactness score (H), inflammatory cell infiltrate presence (I), or periosteal reaction (J). Scoring matrix given in supplementary materials. * indicates significant difference (p < 0.05). Scale bar = 500 μm (A,B,C,D) or 100 μm (E).

Figure 6.

MicroCT analysis of bioreactor specimens. Representative reconstructed microCT slices from central sections of untreated AG (A) and untreated XG (B) bioreactors from a similar region as the histological sections in Figure 6. Residual xenograft (C) was calculated using ImageJ. The control groups were bioreactors packed with filling material (either autograft or xenograft) without implantation. An analysis of all mineralized tissue (includes graft and new bone if implanted) was used to determine bone volume/total volume (D), trabecular spacing (E), trabecular thickness (F), and trabecular number (G) measurements for XG control and AG control (n=3 each) and for untreated XG, treated XG, untreated AG, and treated AG (n=6 each). Data represent mean ± standard deviation. Those that do not share the same letter are significantly different (p < 0.05). Scale bar = 2 mm.

Figure 7.

Histological analysis of bioreactor tissue. Low magnification images from untreated AG (A) and untreated XG (B) with high magnification images showing lamellar bone surrounding degrading autograft (C) or immature woven bone adjacent to xenograft (D). Graft coverage (E), tissue maturity score (F), and osteoclast presence (G) were scored individually by three blinded individuals. Fractional depth (H) was measured using image analysis software. Scoring matrix available in supplemental data. Data represent mean ± standard deviation. * indicates significant difference (p < 0.05). Scale bar = 2 mm (A,B) or 100 μm (C,D).

Figure 8.

Mechanical analysis of bioreactor specimens. Compressive modulus (A), compressive strength (B), and pull-out force (C) were calculated for untreated XG, treated XG, untreated AG, and treated AG (n=6/group). Data represent mean ± standard deviation. Those that do not share the same letter are significantly different.