In vivo caprine model for osteomyelitis and evaluation of biofilm-resistant intramedullary nails

Abstract

Bone infection remains a formidable challenge to the medical field. The goal of the current study is to evaluate antibacterial coatings in vitro and to develop a large animal model to assess coated bone implants. A novel coating consisting of titanium oxide and siloxane polymer doped with silver was created by metal-organic methods. The coating was tested in vitro using rapid screening techniques to determine compositions which inhibited Staphylococcus aureus growth, while not affecting osteoblast viability. The coating was then applied to intramedullary nails and evaluated in vivo in a caprine model. In this pilot study, a fracture was created in the tibia of the goat, and Staphylococcus aureus was inoculated directly into the bone canal. The fractures were fixed by either coated (treated) or non-coated intramedullary nails (control) for 5 weeks. Clinical observations as well as microbiology, mechanical, radiology, and histology testing were used to compare the animals. The treated goat was able to walk using all four limbs after 5 weeks, while the control was unwilling to bear weight on the fixed leg. These results suggest the antimicrobial potential of the hybrid coating and the feasibility of the goat model for antimicrobial coated intramedullary implant evaluation.

Figure 1

Osteoblast proliferation on the coating as a function of silver concentration by weight percent. The results are normalized to the noncoated polystyrene (PS) controls. Data = Mean ± SD.

Figure 2

Staphylococcus aureus viability/proliferation on hybrid surface treatments after one day, as a function of sliver concentration by weight percent. The results are normalized to the noncoated polystyrene (PS) controls. Data = Mean ± SD.

Figure 3

XPS analysis of coated interlocking screw. Silver was present on both unused (a) and used (b) screws. The analysis also confirmed the existence of Ti on the coated surface.

Figure 4

X-ray radiographs of the control (4818) and treated (4820) goat tibias immediately after surgery and 5 weeks after surgery. At each time point, craniocaudal (anteroposterior) and lateromedial views were taken. All images are at the same scale.

Figure 5

Micro-CT images of the tibia after the nails were removed. Images were taken at osteotomy sites and reconstructed using ImageJ. 3D reconstruction, top view and side view images for 4818 control goats ((a), (b), and (c)) and 4820 treated goats ((d), (e), and (f)) are presented.

Figure 6

Composite photomicrographs of H&E stained longitudinal sections from the tibias cortex of control goat (4818) (a) and treated goat (4820) (b). 2X magnification photomicrographs (scale bars = 1 mm) show large periosteal callus (a,b—vertical double-headed arrows) composed of woven bone and fibrous connective tissue surrounding the osteotomy sites, which are not bridged by callus (a,b—horizontal double-headed arrows). Lysis of adjacent cortical bone (a,b—C) is evident, and inflammation (a,b—asterisks) concentrated at the osteotomy site dissects through the cortex of 4818. Folds in the cortex of 4820 (b, white arrows) are artifacts of processing. 20X magnification photomicrographs (insets, scale bars = 50 μm) show necrotic debris (a, inset—asterisk) surrounded by inflammation (a, inset—i) and microsequestra (a, inset—arrow). A large bacteria colony surrounded by radiating arrays of proteinaceous material (b, inset—asterisk), consistent with Splendore-Hoeppli formation is surrounded by inflammation (b, inset—i).

Figure 7

Representative confocal images of biofilm and tissue formed on an IM nail surface. Blue channel (a) and green channel (b) represent DAPI nucleus staining and FITC conjugated S. aureus antibodies. (c) and (d) are overlay images of both channels with 3D projection on x (d1) and y (d2) axes.

Figure 8

DAPI and FITC surface coverage on IM nails explanted from treated goat (4820) and control goat (4818). The surface coverage for DAPI (dark grey) is higher and is expressed using the left y-axis. The surface coverage for FITC (light grey) is showed using the right y-axis. Data = Mean ± SEM. Averages were taken from 10 random 450 μm × 450 μm images.