Less femorotibial rotation and AP translation in deep-dished total knee arthroplasty. An intraoperative kinematic study using navigation.

PURPOSE: Hyper-congruent inserts have been proposed as another means of posterior stabilization in total knee arthroplasty. Their kinematics, partially unexplored, is reported to be possibly erratic. The objectives of the present study were to detect whether prostheses with such a constrained design would provide antero-posterior (AP) stability without interfering with high flexion.
METHODS: The kinematics of 10 knees replaced with hyper-congruent inserts was tested intra-operatively with a specially designed navigation system (Praxim, La Tronche, Isère, France), to measure AP displacements of femoro-tibial contact points at knee flexion.
RESULTS: Femoro-tibial contact points in full extension were in a posterior position compared to their initial position before implantation (8 ± 6 mm medially, 15 ± 10 mm laterally, P < 0.004). AP displacements were different from pre-operative displacements (3 ± 4 mm vs -5 ± 2 mm for the medial condyle, P < 0.01 and -2 ± 6 mm vs -22 ± 8 mm for the lateral condyle, P < 0.001). Forward rolling persisted in four cases, with the medial condyle being involved in three of them (9, 9, and 6 mm, respectively). Post-operative flexion of 122° on average was not correlated with AP displacements.
CONCLUSIONS: Hyper-congruent prostheses partially stabilized femoral condyles at knee flexion. Posterior displacements were reduced with no consequence on range of flexion. Posterior stabilization was imperfect, and paradoxical displacements were detected by navigation, which could therefore help optimize knee balance. LEVEL OF EVIDENCE: Diagnostic study, Level II.