Infection remains a difficult-to-treat complication of total knee arthroplasty. The gold standard treatment is two-stage removal of the prosthesis with later replacement of permanent implants. The first stage consists of removal of the infected arthroplasty components and the surrounding devitalized tissue, copious pulsed irrigation, and placement of a temporary antibiotic-impregnated cement spacer. This spacer typically is left in place six weeks, during which time the patient receives intravenous antibiotics. After the surgeon feels that the infection has been eradicated, or if the patient requires repeat debridement, a second operative procedure is performed. While the use of an antibiotic-loaded spacer is well accepted, whether the spacer should immobilize the knee (a so-called "static" spacer) or allow for range of motion (a so-called "articulating" spacer) is controversial. Proponents of articulating spacers argue that they prevent scarring of the musculature surrounding the knee resulting in easier reimplantation, improved long-term knee function, and improved range of motion. Proponents of static spacers argue that immobilization of the periarticular soft tissues aids in clearance of the infection and is simpler to fashion intraoperatively. While good results have been described with both methods, comparative trials have been conflicting as to whether spacer design alters knee function, operative time, and range of motion. Equipoise exists within the literature, and no randomized clinical trial has been conducted to evaluate this issue.

The purpose of this study is to compare articulating and static antibiotic-impregnated spacers for the treatment of chronic periprosthetic infection complicating total knee arthroplasty through a prospective, randomized clinical trial. The goals of this trial are to determine the effect of spacer design upon eradication of infection, knee function, ease of reimplantation, and range of motion. The investigators hypothesize that articulating spacers will provide shorter operative times at reimplantation, while improving knee function and range of motion.

After diagnosis of infection and informed consent, patients will be taken to the operating room. After anesthetization, patients will be randomized to either an articulating spacer or a static spacer. Randomization will be performed by prepared opaque envelopes administered by a nonparticipant in the study. After a complete debridement of devitalized tissue, explantation of the infected components and any associated cement, either an articulating or static spacer will be placed. All spacers will be formed of 3 g of Vancomycin and 1 g of Tobramycin for each 40 g packet of cement. Articulating spacers will be formed of antibiotic impregnated cement using the Stage One system (Biomet, Warsaw, IN). Static spacers will be hand-made to fit the femoral and tibial exposed metaphyses as a solid block with associated antibiotic cement coated tibial and femoral intramedullary rod, such that knee motion will be minimized.

Post-operatively, all patients will be made touch-down weight bearing protected with a walker or crutches. If a static spacer is placed, patients will be immobilized using with a knee immobilizer. If an articulating spacer is utilized, range of motion will be allowed to the limits of stability as determined in the operating room and protected with a hinged knee brace. At the time of reoperation, the joint will be aspirated and multiple cultures obtained along with intraoperative histopathological analysis to evaluate for persistent infection.

Data collected preoperatively will include age, gender, laterality, etiology of knee degeneration, comorbidities, Knee Society score, and infecting organisms. The Knee Society score has been used extensively in the study of revision knee arthroplasty and has been found to be reliable and valid and will be determined pre-operatively and at all follow-up visits.

Data collected at the time of implant removal and reimplantation will include operative time, blood loss, and need for an extensile exposure. Radiographs performed immediately following and just prior to reimplantation will be reviewed to determine if the spacer utilized has caused bone loss; bone loss to the cut bony surfaces will be confirmed intraoperatively. At each follow-up visit radiographic appearance, the Knee Society Score, knee range of motion, recurrence of infection, and the need for revision or reoperation of any kind on the knee will be determined.

All portions of this study will be part of conventional care except for randomization to either a static or articulating spacer. Which type of spacer is used currently depends upon the judgment of the attending surgeons and both are used routinely.

The primary outcome variable will be range of motion. A power analysis was conducted with the assistance of Dr. Mario Moric at Rush using range of motion as our primary outcome variable, with standard deviations culled from two of the largest series to date -Van Thiel and colleagues (2010) and Fehring and colleagues (2000). For an 80% chance of detecting a predetermined clinically significant difference of 10 degrees, 53 patients per group, 106 patients total, will be needed. To account for attrition, our target sample size will be 140 patients.

Inclusion Criteria:

1) Diagnosis of a periprosthetic joint infection of a primary total knee arthroplasty with a planned two-stage exchange procedure.

Exclusion Criteria:

1. Infection of a revision as opposed to a primary total knee arthroplasty
2. Medically unfit for operative intervention
3. Extensive bone loss preventing the use of an articulating spacer
4. Soft tissue defects that prevent the use of an articulating spacer
5. Known allergy to polymethylmethacrylate, tobramycin or vancomycin.