Despite high success rates following total knee arthroplasty (TKA), knee kinematics are altered following TKA. Additionally, many patients report that their reconstructed knee does not feel ‘normal’ [1], potentially due to the absence of the anterior cruciate ligament (ACL), an important knee stabilizer and proprioceptive mechanism. ACL-retaining implants have been introduced with the aim of replicating native knee kinematics, however, there has yet to be a detailed comparison between knee kinematics in the native knee and one reconstructed with an ACL-retaining implant.

Six fresh-frozen right legs (77±10 yr, 5 male) were mounted in a kinematic rig and subjected to squatting (40°-105°) motions. The vertical positon of the hip was manipulated with a linear actuator to induce knee flexion while the quadriceps were loaded with an actuator to maintain a vertical load of 90 N at the ankle [2]. Medial/lateral hamstring forces were applied with 50 N load springs. During testing, an infrared camera system recorded the trajectories of spherical markers rigidly attached to the femur and tibia. Two trials were performed per specimen. Following testing on the native knee, specimens were implanted with an ACL-retaining TKA (Vanguard XP, Zimmer Biomet) and all trials were repeated. Three inlay thicknesses were tested to simulate optimal balancing as well as over- (1 mm thicker) and understuffing (1 mm thinner) relative to the optimal thickness.

Pre-operative computed tomography scans allowed identification of bony landmarks and marker orientation, which were used define anatomically relevant coordinate systems. The recorded marker trajectories were transformed to anatomical translations/rotations and resampled at increments of 1° of knee flexion. Translations of the medial and lateral femoral condyle centers were scaled to maximum anterior-posterior (AP) width of the medial and lateral tibial plateau, respectively. For all kinematics, statistical analysis between knee conditions was conducted using repeated measures ANOVA in increments of 10° knee flexion.

Internal rotation of the tibia was significantly lower (p<0.05) for the three reconstructed conditions relative to the native knee at flexion angles of 60° and below. No significant differences in tibial rotation were observed between the balanced, overstuffed, or understuffed conditions. The varus orientation was not significantly influenced by implantation, regardless of inlay thickness, for all flexion angles. At 40° flexion, the AP position of the femoral medial condyle was significantly more anterior for the native knee relative to the balanced and understuffed conditions. This finding was not significant for the other flexion angles. No significant differences were found for the lateral condyle center AP position at any flexion angle.

Preservation of the cruciate ligaments during total knee arthroplasty may allow better physiologic representation of knee kinematics. The implants tested in this study were able to replicate kinematics of the native knee, except for tibial rotation and AP position of the medial femoral condyle in early knee flexion. Interestingly, the impact of inlay thickness was generally small, suggesting some tolerance in the choice of inlay thickness.