[Two stages reimplantation for infection after knee arthroplasty. Apropos of a series of 29 cases].

PURPOSE OF THE STUDY: The purpose of this work was to precise diagnosis and treatment of infected total knee arthroplasty with two stage reimplantation.

MATERIAL: 29 infected total knee arthroplasties were operated between 1984 and 1994 and included in this study (mean F.U. 3.5 Y). There were 20 females and 9 males, mean age 70 (46-83). The original arthroplasty was done for OA in 28 patients, RA in one. The arthroplasties were: UHK 2, Bicompartmental 2, Tricompartmental 19. 20 TKA were cementless. 14 patients showed one or several risk factors. Infection was diagnosed in 1 of 2 ways: preoperative aspiration or culture of surgical specimen. There were 12 staphylococcus epidermidis, 8 staphylococcus aureus, streptococcus (n = 2) acinetobacter (n = 2), peptococcus (n = 1) pseudomonas (n = 1), gemella morbidellum (n = 1). 6 were non identified.

METHOD: The protocol for two stage reimplantation began with components and cement removal. A synovectomy was performed. The knee cavity was filled with antibiotic cement spacer and the wound was closed. The leg was placed in a splint. All patients underwend a continued antibiotic therapy, specific in 20 cases with isolated organisms. A total knee arthroplasty was performed, using a total posterior cruciate substituting prosthesis, 6 to 8 weeks after components removal (2-24). All patients received parenteral antibiotics after reimplantation for not less than 2 months (2-6).

RESULTS: Infection was eradicated in 24 cases, 22 in one time, 2 bad second debridement. At last follow-up the average Hungerford score was 75.6/100, the average Knee society knee score was 80 and the average functional score was 70. Mean range of flexion was 95 degrees. 6 patients had recurrent infection and poor result. They underwent arthrodesis. 5 of the 6 patients had solid mature fusion at last follow-up.

DISCUSSION: The results of two stage reimplantation for infected total knee replacement showed that this is the method of choice for infection treatment and acceptable function restoration. As other authors, we get a good success rate (82 per cent). Functional result was better with identified microorganisms, but we did not find any correlation with organisms type or infection length. Punction and bone scanning are of great help for diagnosis in difficult chronical cases. Organism identification is fundamental for infection duration. Staphylococcus epidermidis was the most frequent identified organism. New procedures using articulated cement spacer may improve functional results.