Methodology: A retrospective review based on a prospective database was performed on 146 consecutive revision TKA’s. An independent observer measured clinical outcomes using the Knee Society Knee (KS) and Function Score (FS). X-ray evaluation, including rating of radiolucent lines, tibiofemoral and patellofemoral alignment, was carried out by an independent radiologist. ANOVA was used for statistical analysis, with significance set at p≤0.05 (SPSS version 15.0). Post-hoc Bonferroni testing was carried out for single variables including primary cause of failure, age at revision surgery, time span between index operation and revision, type of index operation, partial or total revision and the performance of a tuberosity osteotomy.
Results: 146 files were available in 135 patients. 16 patients deceased (17 knees) during the follow-up period and 2 patients (2 knees) were lost to follow-up. 117 patients (127 knees) were available for evaluation. Age at revision surgery averaged 67.7 years (range 32.3–88.1). Mean follow-up time was 4.5 years (range 1–14). Patients had revision TKA between 51 days and 16.1 years (average 4.7 years) after the index TKA. 54% of the early revisions were due to infection and instability, 55% of late revisions were caused by polyethylene-wear and loosening. The mean postoperative KS was 70.8 with a mean improvement of 43.2 points as compared to pre-operative. The mean postoperative FS was 52.9 with a mean improvement of 25.4 points. Grouping outcomes according to cause of failure of the index TKA gave the following ranking from better to worse, without being significant: wear (n=15; KS 80.8; range 43–99, SD 17.5), loosening (n=44; KS 75.8; range 15–100, SD=21.2), malalignment (n=19; KS 70.0; range 9–95, SD 25.9), instability (n=33; KS 68.2; range 5–100, SD 24.1), others (n=16; KS 66.7; range 10–100, SD 25.9), and infection (n=21; KS 64.2; range 3–100, SD 31.7). Survivorship at 5 years was 90.0% (CI 86.4% –93.6%), at 10 years 84,6% (CI 77.0% –92.3%) and at 14 years 84,6% (CI 37.7% –131.6%). Significant better outcomes were seen with late revisions, index operation being partial knee replacement and older age at revision. More failures (p=0.002) were seen with early revisions. In 32.6% of the patients radiolucent lines of ≥1 mm were observed. Points were granted with the use of a Radiolucency Scoring Scheme. Patients with less than 4 points (n=87, mean KS 71.2) had better outcomes than patients with 4 or more points (n=8, mean KS 56.4). 87% of patients were aligned within 4° of mechanical axis.
Conclusion:
Outcomes of revision TKA are inferior to primary TKA.
Early failures were mainly caused by infection, instability, malalignment.
Grouping revision TKA’s to etiology of failure did not lead to significant differences in outcomes.
Significant better outcomes were reported for late revisions, patients with older age at revision surgery and partial knee replacement.
Survivorship analysis was significally better for late than for early revisions.
