[Revision total knee arthroplasty after unicompartmental femorotibial prosthesis: 54 cases]

We analyzed technical difficulties encountered when performing revision total knee arthroplasty in patients with unicompartmental femorotibial prostheses.

This multicentric retrospective study included 54 revisions of unicompartmental femorotibial prosthesis with implantation of a total knee prosthesis. The series included 45 medial and nine lateral compartment prostheses. A gliding total knee prosthesis was implanted in 53 cases (98%) (39 standard, 14 revision). Mean time to failure of the unicompartmental prosthesis was four years. IKS scores were established at review. The radiological work-up included AP and lateral views in single leg stance and goniometry for 22 medial compartment revisions. Twenty-seven patients were seen for physical examination and x-rays and eight were lost to follow-up; data were recorded from medical files for 19 patients.

The revision procedure was considered easy in 82% of the cases. Mean follow-up after revision was four years (range 2 - 12 years). Subjective outcome was very satisfactory for 56% of the patients, satisfactory for 36% and unsatisfactory for 8%. The mean function score was 62 points, the mean knee score 85 points, and the mean flexion was 113 degrees. No laxity was found for 90% of the knees. The femorotibial angle was 180 +/- 2 degrees in 46% of the patients. The mechanical femoral angle was 90 degrees in 54% of the patients with 2-4 degrees varus in 42%. The mechanical tibial angle was 90 degrees in 46% of the patients with 2-8 degrees valgus in 37%. Complications included pulmonary embolism (n=2), mobilization under general anesthesia (n=3), arthrolysis (n=1), lateral vertical patellectomy (n=1), and secondary infection (n=1). There were five failures requiring changing the total knee prosthesis.

Loss of bone stock raises specific problems during revision of unicompartmental knee prostheses. Loss of tibial bone is more frequent but it is more difficult to correct for loss of femoral bone. A gliding knee prosthesis is generally preferred for first intention revision. We recommend a long stem when the bone defect is important or involves loss of cortical bone. We have had good mid-term results with revision total knee prostheses after unicompartmental prostheses. Longer follow-up is needed. Poor results were obtained when revision was performed for persistent pain without a clearly defined cause. The presence or not of significant bone loss did not appear to affect outcome. The observation of medial laxity in case of failed lateral unicompartmental prostheses suggests a more constrained total knee prosthesis might be indicated. Compared with earlier series, our results with total knee prostheses after unicompartmental prostheses appear to be better than after tibial valgus osteotomy and also better than after total knee arthroplasty. Conversely, they would be less satisfactory than for primary total knee arthroplasty. The surgical procedure for revision total knee arthroplasty after unicompartmental prosthesis requires precision and skill but is not technically difficult.