Author(s): Jay M. Zampini, MD, Gwo-Chin Lee, MD, and Norman A. Johanson, MD; Drexel University College of Medicine, Philadelphia, PA
Title: An Analysis of Factors Contributing to Failure of Two-Stage Procedures for Revision of Infected Total Knee Arthroplasty
Purpose: The purpose of this study is to evaluate the host, surgical, and perioperative factors leading to failure of two-stage procedures for chronically infected total knee arthroplasty.
Methods: A retrospective search of the hospital medical record archives and the operating room patient data base was performed to identify all patients diagnosed with a chronic, deep infection of a total knee replacement. A deep, chronic infection was defined by clinical evaluation, elevated C-reactive protein, erythrocyte sedimentation rate, peripheral white blood cell count, and/or growth of bacteria from a joint aspirate in patients symptomatic for four or more weeks. Each patient underwent resection arthroplasty with placement of an antibiotic loaded cement spacer. Culture specific antibiotics were administered intravenously for a minimum of six weeks followed by confirmation of clinical success by normalization of inflammatory lab values and no bacterial growth from a joint aspirate. Reimplantation of a total knee replacement was performed only when indicated. A successful procedure was defined as reimplantation with no recurrence of infection or need for chronic suppressi! ve antibiotic treatment. We hypothesized that host, surgical, and perioperative factors influence the ability to eradiate infection in chronically infected total knee replacements. Each case was evaluated to determine the variables contributing to failure of reimplantation. Statistical comparison was made using Student's t-test, Chisquare analysis, and Fisher's exact test where appropriate.
Results: Between January 1, 1998 and April 30, 2006, thirty patients underwent two-stage revision procedures for infected total knee arthroplasty. Fifteen were treated successfully without recurrence of infection. Eleven patients failed two stage reimplantation procedures and four were lost to follow-up. There were 18 men and 12 women with a mean age of 57 years. Both groups were statistically similar with respect to age, gender, and number of comorbidities. The mean number of operations prior to explantation was 4.7 in the failure group compared to 2.6 in the success group (P = 0.035). Diabetes occurred 5 times more commonly in the failure group (P = 0.041) and allogeneic blood transfusion occurred 6 times more commonly in the failure group(P = 0.034). The most common infecting organism was methacillin resistant Staphylococcus aureus. Polymicrobial infections were discovered in four failed an in no successful revisions (P = 0.029). Review of radiographs revealed retained polymethy! l methacrylate cement in seven failed cases and in no successful cases (P = 0.009). Despite multiple debridements and cement spacers, persistent infection led to arthrodesis in four knees and transfemoral amputation in three.
Discussion: Many factors are known or purported to affect the ability to eradicate deep, chronic infection in total knee replacement. The results confirm that systemic and local host factors (diabetes and multiple knee operations, respectively) diminish that ability to eradicate infection. Similarly, the immune modulation resulting from allogeneic blood transfusion is shown here to contribute to reinfection. These results also confirm that perioperative identification of a polymicrobial infection greatly decreases the chances of eradicating infection. Finally, the influence of retained cement after initial debridement strongly reinforces the importance of the adequacy of the debridement to achieve success in clearing deep, chronic infection in total knee replacements.