Use of allograft bone has become standard for bridging defects unlikely to heal by simple fixation and routinely used in revision arthroplasties for implant stabilization. Unfortunately, this decellularized allograft provides an ideal surface for bacterial colonization, necessitating repeated surgeries, extensive debridement and lengthy antibiotic treatments. With up to 18% infection rate following allograft surgeries, a need for more effective means to prevent this process is evident. We describe a novel modification of native bone allografts that renders their surface bactericidal while increasing the effectiveness of systemic antibiotic treatments.

Allograft modification: Morselized human bone was washed extensively and sequentially coupled: 2X with Fmoc-aminoethoxyethoxyacetate (Fmoc-AEEA); deprotected with 20% piperidine in Dimethylformamide (DMF); and then coupled with vancomycin (VAN) for 12–16 hours. The VAN-bone was washed extensively with DMF and PBS for at least 1 day. VAN immuno-fluorescence: Control or VAN-bone was washed 5X with PBS, blocked with 10% FBS (1hr), incubated with rabbit anti-VAN IgG (4oC, 12h) followed by an Alexa-Fluor 488-coupled goat anti-rabbit IgG (1hr), and visualized by confocal laser microscopy. Antibiotic Activity. Equal dry weights of control and VANbone were sterilized with 70% ethanol, rinsed with PBS, and incubated with either Staphylococcus aureus (S. aureus) or Escherichia Coli (Ci=104 cfu) in TSB, 37oC, for 2, 5, 8 and 12 hrs. Antibiotic treatment: Clinical grade vancomycin was added to the solution with bacteria or following infection at a final concentration of 10μg/ml. Bacterial counts: Non-adherent bacteria were removed by washing and adherent bacteria suspended by sonication in 0.3% Tween-80 for 10mins followed by plating on 3M^® Petrifilms. Bacterial visualization: Non-adherent bacteria were removed by washing extensively with PBS and adherent bacteria stained with the Live/Dead BacLight Kit (20mins, RT) to cause viable bacteria to fluoresce green. Samples were visualized by confocal microscopy.

In comparison to controls, VAN-bone consistently reduced the graft bacterial load by ~90% at all time points. After staining and visualization of adherent bacteria, biofilm formation was apparent on controls by 12 hrs and absent from VAN-bone. E.coli, a gramnegative organism that is not sensitive to VAN, readily colonized both control and VANbone, confirming retention of VAN specificity. We then evaluated VAN-bone activity in a system that modeled systemic antibiotic therapy and antibiotic prophylaxis. In the absence of solution antibiotics, VAN-bone exhibited a significant decrease in bacterial colonization as compared to controls. When 10 μg/ml VAN was added to the medium for the last 4 h (modeling systemic antibiotic therapy), colonization of control surfaces was reduced, while colonization of VAN-allograft was almost eliminated. When 10 μg/ml VAN was added concomitantly with S. aureus, VAN-bone colonization was undetectable, while colonization of control surfaces still occurred.

We have previously described an antibiotic-tethered allograft that resists bacterial colonization. In this abstract, we test this technology with an vitro model of bone implantation in the presence of solution antibiotics. In these models, solution antibiotics failed to prevent infection of control bone while completely clearing the bacteria on VAN-bone. Furthermore, VAN bone exhibited high activity against S. aureus, a gram positive organism, whereas it was ineffective against E. coli, a gram negative organism. The specificity of the tethered antibiotic supported the view that the antibacterial properties of the allograft were related to the tethered antibiotic and not to undefined aspects of the attachment chemistry. In terms of antibacterial activity, when challenged with 104 CFU S. aureus (with concentrations reaching > 107 CFU by 24 h), the antibiotic -modified allograft consistently decreased bacterial colonization by > 90%; S. aureus inocula < 102 CFU resulted in no detectable colonization of the VAN-allograft. Thus, development of these allografts may not only combat allograft colonization but increase the effectiveness of prophylactic antibiotics to ultimately result in a new therapy for allograft-associated infection.

INTRODUCTION: We have previously demonstrated the efficacy of a modified Ti-surface tethered with antibiotics in preventing bacterial colonization. It is not known if coverage of this surface with serum or other physiological material may hinder the bactericidal properties of such a surface.. The in vitro activity and efficacy of such a surface against S. aureus and S. epidermidis was tested following coverage of the surface with serum.

METHODS: Vancomycin was coupled to Ti6Al4V pins by aminopropylation, linker addition, and vancomycin coupling (VancTi). Bactericidal activity was tested in solutions of bacteria (Ci=1×104cfu/ml) incubated with pins±pre-incubation with fetal bovine serum (FBS). Anti fibronectin and anti-vancomycin antibodies were used to detect surface coating or modification. Bacteria were detected by fluorescent labeling (Syto9) or by direct counting after solubilization.

RESULTS: By immunofluorescence, pins were extensively covered with serum fibronectin which did not interfere with the diffuse, intense vancomycin staining. When incubated with S. epidermidis or S. aureus, VancTi showed little colonization compared to control pins (> 95% reduction in cfu).

DISCUSSION: In a physiological environment, implants will be coated with serum proteins. Activity of the VancTi was unaffected by this coating and maintained potent inhibition of bacterial colonization. We have described a surface modification that allows Ti implants to resist colonization and subsequent periprosthetic infection. Such surfaces hold great promise for the prevention and treatment of periprosthetic infections.

Introduction: We have previously described modification of Ti that renders the implant surface bactericidal and prevents bacterial colonization in vitro. This study evaluates the efficacy of the same surface in preventing periprosthetic infection in a small mammal model.

Methods: Control or VancTi rods were incubated with S. aureus (Ci=104CFU/mL) in TSB containing 0, 5×10-3, 4, or 100 μg/mL vancomycin at 37°C for 24h. Bacteria were detected by fluorescence (Live/Dead BacLight) and imaged by confocal microscopy. Resistance was tested by incubating control or VancTi rods with S. aureus for 0–8 weeks. Adherent bacteria were tested every 7 days on vancomycin screening agar (6μg/mL).

Results: Using a percutanous approach, the intercondylar region of the knee in rats were identified. The intrameduallry canal of the femur was retrograde reamed using an 18-gauge needle. Infection was induced by injection of 103 CFU S. aureus in 150uL of saline into the femoral canal followed by insertion of a 20 mm Ti implant. Animals were harvested at various time points. At harvest, animals were euthanized with CO2.. Detailed analysis including radiographic, micro-CT, histological, bacteriological, and clinical evaluation was performed.

Results: All animals showed signs of infection within the first few post-operative days with increased soft tissue swelling and limited mobility. At 1 week 75% of the animals showed radiographic signs of periprosthetic infection including periosteal reaction, abscess formation, widened canal, bone destruction, and formation of involucrum. PPI could be prevented in 92% of cases when modified Ti-Van was used. In one animal despite the use of antibiotic-tethered implant, PPI occurred partially due to the pin insertion penetrating the bone cortex.

Discussion: Biologically modified implants with bactericidal surface can have a promising role in management of periprosthetic infection. The modified implant described herein contains a nanoscale surface of covalently linked antibiotics that can potentially confer bactericidal properties throughout the life of the implant