Femoral Component External Rotation Affects Knee Biomechanics: A Computational Model of Posterior-stabilized TKA

Background: The correct amount of external rotation of the femoral component during TKA is controversial because the resulting changes in biomechanical knee function associated with varying degrees of femoral component rotation are not well understood. We addressed this question using a computational model, which allowed us to isolate the biomechanical impact of geometric factors including bony shapes, location of ligament insertions, and implant size across three different knees after posterior-stabilized (PS) TKA.

Questions/purposes: Using a computational model of the tibiofemoral joint, we asked: (1) Does external rotation unload the medial collateral ligament (MCL) and what is the effect on lateral collateral ligament tension? (2) How does external rotation alter tibiofemoral contact loads and kinematics? (3) Does 3° external rotation relative to the posterior condylar axis align the component to the surgical transepicondylar axis (sTEA) and what anatomic factors of the femoral condyle explain variations in maximum MCL tension among knees?

Methods: We incorporated a PS TKA into a previously developed computational knee model applied to three neutrally aligned, nonarthritic, male cadaveric knees. The computational knee model was previously shown to corroborate coupled motions and ligament loading patterns of the native knee through a range of flexion. Implant geometries were virtually installed using hip-to-ankle CT scans through measured resection and anterior referencing surgical techniques. Collateral ligament properties were standardized across each knee model by defining stiffness and slack lengths based on the healthy population. The femoral component was externally rotated from 0° to 9° relative to the posterior condylar axis in 3° increments. At each increment, the knee was flexed under 500 N compression from 0° to 90° simulating an intraoperative examination. The computational model predicted collateral ligament forces, compartmental contact forces, and tibiofemoral internal/external and varus-valgus rotation through the flexion range.

Results: The computational model predicted that femoral component external rotation relative to the posterior condylar axis unloads the MCL and the medial compartment; however, these effects were inconsistent from knee to knee. When the femoral component was externally rotated by 9° rather than 0° in knees one, two, and three, the maximum force carried by the MCL decreased a respective 55, 88, and 297 N; the medial contact forces decreased at most a respective 90, 190, and 570 N; external tibial rotation in early flexion increased by a respective 4.6°, 1.1°, and 3.3°; and varus angulation of the tibia relative to the femur in late flexion increased by 8.4°, 8.0°, and 7.9°, respectively. With 3° of femoral component external rotation relative to the posterior condylar axis, the femoral component was still externally rotated by up to 2.7° relative to the sTEA in these three neutrally aligned knees. Variations in MCL force from knee to knee with 3° of femoral component external rotation were related to the ratio of the distances from the femoral insertion of the MCL to the posterior and distal cuts of the implant; the closer this ratio was to 1, the more uniform were the MCL tensions from 0° to 90° flexion.

Conclusions: A larger ratio of distances from the femoral insertion of the MCL to the posterior and distal cuts may cause clinically relevant increases in both MCL tension and compartmental contact forces.

Clinical Relevance: To obtain more consistent ligament tensions through flexion, it may be important to locate the posterior and distal aspects of the femoral component with respect to the proximal insertion of the MCL such that a ratio of 1 is achieved.