There is ongoing debate on the benefits of fixed versus mobile bearing Unicompartmental Knee Replacement (UKR). We report the results from a randomised controlled trial comparing fixed and mobile bearing of the same UKR prosthesis. Forty patients were randomized to receive identical femoral components and either a fixed or mobile bearing tibial component. At 6.5 years follow-up 37% of the mobile bearing design had been revised and 14% for the fixed bearing design. The main reasons for revision were pain and loosening. These results were compared with data from The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) that show a cumulative percent revision of 24.2% for the mobile bearing Preservation UKR at 6.5 years. All locally explanted mobile bearings were examined microscopically, and 83% demonstrated significant backside wear. Constraint on the undersurface of the bearing coupled with a congruent upper surface may have contributed to the excessive revision rate. This is the first randomised controlled trial examining mobile and fixed variations of the same UKR prosthesis and shows this design of UKR with the mobile bearing has an unacceptably high revision rate and patients with this knee design should be closely monitored.

Introduction An approximation of normal knee kinematics after knee replacement may improve knee function and implant fixation and reduce wear of the prosthesis. This study describes the knee joint kinematics after unicondylar knee arthroplasty (UKA) in general, and compares the Miller-Glante (MG, fixed bearing) and Oxford (mobile bearing) implants in particular.

Methods Twenty-two knees in 17 patients (11 males, six females, mean age of 69.7 yrars) were randomized into MG (11 knees) or Oxford (11 knees). No clinical complications or signs of loosening were observed. At the one year follow-up, RSA (Radiosterometry) x-rays were taken by using two x-ray tubes positioned at knee level and exposing the knee simultaneously from the side. Four pairs of weight bearing x-ray were obtained at zero degrees, 30°, 60°, 90° of knee flexion, with zero as reference position. Tibial rotation, rollback, translation of tibia-femur contact point, and the bearing movement were analyzed using UmRSA software.

Results With the MG implant, the tibia internally rotated 3.0°, 3.0°, and 4.2° respectively at 30°, 60°, and 90° of flexion, while with the Oxford implant, the tibia internally rotated 4.3°, 7.6°, and 9.5° respectively at 30°, 60°, and 90°. No significant difference was found between the two groups (P> 0.05, Repeated-measures ANOVA). The medial femoral condyle moved backward (1.8 and 1.5 mm respectively in MG and Oxford) from zero degrees to 30° of flexion. At 60°, it moved anteriorly in both knees, in MG to 0.9 mm anteriorly and in Oxford to 0.6 mm posteriorly to the reference position. At 90° the condyle moved 4.2 mm (MG) and 0.7 mm (Oxford) anteriorly to the reference position. No significant difference between the groups (P> 0.05). The femur-tibia contact point in MG moved anteriorly 2.8, 5.1, and 3.9 mm, respectively at 30°, 60°, and 90° of flexion, whereas the contact point in Oxford moved posteriorly 2.6, 1.8, 2.4 mm respectively at 30°, 60°, and 90°. A significant difference was found between the groups (P=0.003). The bearing in the Oxford implant moved backward of 2.2, 2.0, and 0.9 mm respectively at 30°, 60°, and 90° of knee flexion.

Conclusions The in-vivo weight bearing 3D knee kinematics after UKA with fixed or mobile bearing was described. In MG the medial femoral condyle moved forward with knee flexion, whereas in Oxford it moved backward together with the bearing, which is closer to normal knee kinematics.