Introduction: In general a loose-balanced total knee arthroplasty (TKA) are believed to gain good postoperative range of motion (ROM), however, too much laxity is thought to be a cause for persistent pain and catastrophic long-term results.

Materials and methods: We measured the antero-posterior and medio-lateral laxity to evaluate the influence of stability after cruciate-retaining TKA on ROM, pain score in Knee Society score and Functional score at 4–7 years after the operation. Twenty-one knees in 15 patients with an average age of 68 years (range, 58–78) who had a PCL retaining TKA for osteoarthrosis were examined (YS-4, 16 knees; AGC-S, 5 knees). There were 12 women and 3 men. Anteroposterior laxity was examined using a KT 2000 arthrometer at 30-degree knee flexion. Mediolateral laxity was examined at full extension with manual stress x-rays.

Results: Average ROM was 112 degrees (range, 90–140 degrees), antero-posterior laxity was 9.7 mm (range, 2–27 mm) and mediolateral laxity was 10.6 degrees (range, 5–22 degrees). Knee Society pain score was 47 (range, 10–50) and Functional score was 82 (range, 40–100). No significant difference in ROM and clinical scores were noted between the lax and the stable knees. Correlations between ROM and anteroposterior or medio-lateral laxity were not significant (P = 0.55, P = 0.05, respectively; Spearman’s correlation efficient).

Discussion and conclusion: A loose-balanced TKA did not gain good post-operative ROM. No parameters suggested that lax knees got better flexion angle, worse pain score and functional score than stable knees.

Although bone loss and ligamentous instability are usually indications for the use of constraining prostheses in revision total knee arthroplasty (TKA), several reports have documented a high rate of failure with these prostheses. We therefore tried using the cruciate retaining augmentable type prosthesis (NexGen CRA) if a good intraoperative ligamentous balance could be obtained with revision TKA.

CRA was used on nine knees of seven patients with an average age of 71.7 years. The follow-up lasted for an average of 1.6 years. Clinical evaluation consisted of instability ratings, knee score, range of motion and %MA (mechanical axis) as an index of the alignment. These patients showed improvement varus instability from a mean of 7.0° preoperatively to a mean of 1.5° postoperatively, and in valgus instability from 5.3° to 1.3°.

The knee score was significantly improved from 41 to 81, but the %MA resulted in an unsatisfactory improvement from −8.8% to 34.0%. The femoral anatomical-mechanical angle (FAMA) was measured as a parameter of bowing deformity of the femur, and the knees were divided into two groups, the bowing group consisting of the knees with an FMA of more than 8° (n=6, average 9.0°), the normal group of knees with an FAMA of less than 7° in FAMA (n=3, average 6.7°). The postoperative %MA showed a mean value of 40.9% for the normal group, and 30.6% in the bowing group. These results demonstrated that the discrepancy between FAMA for the bowing group(=9.0°) and the valgus angle of the stem of the femoral component (6.0°) was the cause of the malalignment in the bowing group. These clinical results suggest that the cruciate retaining augmentable type prosthesis can be used successfully for selected revision cases, but that malalignment in knees with bowed femora may remain a problem.