Knee arthrodesis in the treatment of failed total knee replacement.

Twenty-six primary arthrodeses and 6 repeat arthrodeses were done for a total of 32 arthrodesis procedures in 26 patients, with followup from 2 to 10 years (mean, 4 years). The indications for arthrodesis were septic failure in 18 patients and aseptic loosening in 8 patients. Thirteen arthrodeses were done using external fixators, and 12 arthrodeses were done using a closed fluted intramedullary nail for fixation (1 arthrodesis was done with a custom proximal and distal interlocked nail). In addition, 6 repeat arthrodeses with intramedullary nail fixation were done for treatment of nonunions after external fixation. Five (38%) of the 13 patients who underwent arthrodesis with an external fixator had clinical and radiographic union at a mean of 5 months. All 13 patients with primary intramedullary nail arthrodesis achieved union. The knees of the 6 patients with nonunions after external fixation that were treated with repeat intramedullary nail arthrodesis achieved union. Patients with septic failure had staged debridements before intramedullary nail arthrodesis. Intramedullary nail arthrodesis can be done safely in patients with sepsis as a staged procedure. Knee arthrodesis using intramedullary nail fixation gives a much higher union rate than does external fixation and is associated with fewer complications.