Effect of the kinematic retaining design on knee kinematics in total knee arthroplasty: A cadaveric study using a navigation system.

Background: Implant design in total knee arthroplasty (TKA) has evolved considerably, with recent developments focusing on reproducing native knee kinematics. Some implants now feature anatomically and physiologically accurate articular surface geometries. This study aimed to evaluate the impact of different implant designs on knee kinematics using the same cadaveric specimens to ensure consistent comparison. We hypothesized that implant designs incorporating features intended to replicate native joint anatomy, such as the kinematic retaining (KR) design, would more closely reproduce physiological knee kinematics.

Methods: TKA was performed on nine Thiel-embalmed cadaveric knees with mild medial osteoarthritis, using three implant designs from the Physica system: KR, cruciate retaining (CR), and medial congruent (MC) designs. All procedures were performed using a mechanical alignment technique, with both the posterior tibial slope and femoral rotational angle standardized at 3°. The posterior cruciate ligament was preserved throughout the evaluation of all implant designs. A navigation system was used to collect detailed kinematic data. Evaluations were conducted after trial component placement, focusing on anteroposterior, mediolateral, and compression-distraction positions, as well as rotational angles. From these knee status data, femoral rotational kinematics relative to the tibia and the anteroposterior translation of both femoral condyles during flexion were also calculated.

Results: No significant differences in flexion and extension angles were observed between the groups. The KR group presented the greatest mean femoral external rotation relative to the tibia throughout the range of motion among the groups; however, there were no statistically significant differences. The CR and MC group showed significantly reduced anteroposterior translation of the lateral condyle compared with the native knee (p = 0.021 and 0.003, respectively). Furthermore, the anteroposterior translation of the lateral femoral condyle was significantly greater in the KR group than in MC groups (p = 0.021). In the KR group, six of nine knees exhibited medial pivot motion, compared with three in the CR group and four in the MC group.

Conclusions: Using identical cadaveric specimens and navigation-based analysis, we identified distinct kinematic profiles associated with each implant design. Notably, the KR implant demonstrated kinematics approximating native knee motion; however, these findings remain preliminary and warrant further clinical validation.

Level of evidence: III.