Two-stage reimplantation in infected total knee arthroplasty

Twenty-one infected total knee arthroplasties (TKA) in 21 patients were treated from September 1980 through October 1987. Of these, 15 were followed for more than one year. Treatment of these patients consisted of thorough debridement of all infected tissue and components; a cement spacer was used in ten patients. The cement was impregnated with antibiotics. This procedure was followed for an average of 4.2 weeks with intravenous antibiotics and TKA utilizing antibiotic-impregnated cement. Five patients had rheumatoid arthritis and ten had osteoarthritis. The organisms included Staphylococcus coagulase negative (seven patients), Streptococcus group B (two patients), Streptococcus bovis (one patient), Enterococcus (one patient), Staphylococcus coagulase positive and Bacillus circulans (one patient), Staphylococcus coagulase negative and Enterococcus (one patient), Staphylococcus coagulase negative and Pseudomonas aeruginosa (one patient), and Clostridium perfringens (one patient). Of the 15 patients, 12 appeared to be free of infection, two were obvious failures and required knee fusion, and one was suspected of having continued infection at five years and was treated elsewhere. Eleven patients with revision TKA were available for follow-up examinations at an average of 2.9 years (range, one to six years). One patient died five years after reimplantation but had been functioning well. One patient functioning at three years postreimplantation did not return for a later follow-up examination. The average knee score (modification of the Hospital for Special Surgery Knee Score) was 75.5 points (range, 48-94); average flexion was (81° (range, 52°-120°), and average extension was +6° (range, 0°-30°). Roentgenograms were highly suspicious for loosening in one patient with suspected chronic infection. Complications included deep venous thrombosis in four patients, high probability of pulmonary embolus in two patients, superficial wound dehiscence in three patients, greater than six days of serous or hematoma drainage in four patients, and an intraoperative medial femoral condyle fracture in one patient. Two-stage TKA had an overall success rate of 80% and a 90% success rate (not statistically significant) when an antibiotic cement spacer was used.