A Dose-Response Curve for a Gram-Negative Spinal Implant Infection Model in Rabbits
A randomized complete block animal spinal implant infection model with internal control.Objective.
The aim of this study was to develop a spinal implant animal infection model to simulate postoperative gram-negative wound infection.Summary of Background Data.
Implant-associated surgical site infections (SSIs) remain a dreaded complication of spinal surgery. Currently, over 30% of all spine SSIs are secondary to gram-negative bacteria. Traditional animal models have utilized gram-positive inoculums to simulate postoperative infection, but there exists no model in the literature for gram-negative infection in the setting of spinal instrumentation.Methods.
Five New Zealand white female rabbits underwent simulated partial laminectomies and implantation of a 5 mm titanium wire adjacent to the spinous processes of vertebra T4, T9, L1, and L6 to mimic posterior spinal instrumentation. The second site, T9, was used as the sterile internal control sites, while all other sites were challenged with varying inoculums of Escherichia coli (EC American Type Culture Collection 25922): 102, 103, 104, 105, and 106 Colony Forming Units (CFU). The rabbits were sacrificed 4 days postoperatively and bacterial loads were assayed from the implants and surrounding tissue.Results.
No evidence for infection was observed in any of the sterile control sites. The lowest inoculum of E. coli (102 CFU) did not produce a reliable infection. Inoculation with 103 CFU created a consistent soft tissue infection, but inconsistent infection on implants. Inoculation with 105 CFU was required to consistently produce both soft tissue and implant infection.Conclusion.
Consistent soft tissue and implant infection was produced with inoculation of 105 CFU of E. coli. Gram-negative infections represent greater than 30% of all spinal SSIs, and this animal model can reliably reproduce such infections with spinal instrumentation that can guide future development of anti-infective therapies.Conclusion.
Level of Evidence: 2