2006 Abstract : 2- 6

Print Friendly, PDF & Email

Authors: Antoci Jr, Valentin*; Adams, Christopher S; Freeman, Theresa A.; Hickok, Noreen J; Shapiro, Iriving M; Parvizi, Javad

Title: A Rat Model of Prosthesis Associated Infection

Addresses: 1015 Walnut St. 501, Philadelphia, PA19107

Purpose: Few readily usable animal models are currently described, with no models in smaller mammals like the rat. We describe a rat periprosthetic infection model that appears simple and consistent.

Methods: 1mm diameter Ti90Al6V4 (Ti, Goodfellow) rods were passivated, silanized with APTS, reacted twice with AEEA, and covalently linked with Vancomycin. Animal Testing. All protocols were approved by the IACUC of Thomas Jefferson University. Wistar rats (Charles Rivers), 300-350 g, were anesthetized with IP ketamine/xylazine, with maintenance isoflurane, and buprenorphine for pain control bid as needed. Infection was induced by injection of 103, 105, 107 CFU S. aureus in 150uL of saline into the femoral canal with the opposite side without bacteria left as control. Control Ti were sterilized by incubation in 70% ethanol for 15 min, rinsed 3X with PBS, and implanted retrograde in the femoral canal. At harvest, animals were euthanized with CO2, radiographed, femur dissected, microCT. The pin was removed and rolled on plate, followed by sonication and serial dilution plating for CFU counts.

Results: We have tested various methods of inducing bone infection in the rat, including culturing a biofilm on the rod for 72 h, dipping the rod in active bacterial cultures, and injecting the bacteria directly into the femoral canal. Rods with expected biofilm on the surface produced 50% infection in 1 of 2 animals. The most consistent and successful method of inducing periprosthetic infection in the rat seemed to be direct injection into the femoral canal with all tested doses showing good results. The higher 107 CFU/mL causes extensive distress in the animal with extensive swelling and loss of weight bearing. The infection was manifested radiographically within the first 4 days post-operatively. At week 1, the upper two doses showed infection. Upon harvesting of the higher dose group, severe bone loss was observed with extensive puss and fibrous inflammatory tissue. However, no cross over of infection was observed from the infected side to the control side. The low dose group showed mild bone infection, some bone remodeling, with minimum bone loss.

Discussion: Few periprosthetic infection models exist in smaller mammals, and are limited to rabbits, dogs, or sheep. Considering that periprosthetic infection is currently the second most common complication after joint replacement, the need for a good and inexpensive animal model is crucial. We have developed a new model of periprosthetic infection in rats, which are readily available, relatively inexpensive, and easy to manage. However, the rat immune system is highly developed and able to resist severe microbial attack. We have consistently obtained bone infection after injecting directly into the femoral canal as little as 103 CFU in 100 uL of saline. Lower doses have proved unsuccessful. With increasing dose the time line of infection change little, with most drastic differences seen in volume of bone loss and overall bone destruction. At the same time, even with the higher doses the risk of systemic infection and septicemia is very small, and at no time any cross over infection was observed. The proposed rat model of peri-prosthetic infection may provide a ground for more studies to elucidate the mechanisms of peri-prosthetic infection and develop new strategies against implant association infections.