Revision total knee arthroplasty with cemented components and uncemented intramedullary stems.

Sixty-three failed total knee arthroplasties in 60 patients (27 females, 33 males; average age, 66 years) were treated consecutively with revision using cemented component fixation and an uncemented stem. Patients were followed for a mean of 5.75 years (range, 2-10 years); none were lost to follow-up. There were 12 (19%) re-revisions: 6 (10%) were revised for aseptic loosening, 4 (6%) for recurrent infection, and 2 (3%) for instability. Knee Society Pain Scores improved from 56 to 81, and function scores improved from 49 to 62 points. Latest radiographs in retained knees showed none with definite femoral loosening but 4 with tibial component loosening. Combining those revised for aseptic loosening and radiographic aseptic loosening, mechanical failure occurred in 10 patients (16%).