Referencing the center of the femoral head during robotic or computer-navigated primary total knee arthroplasty results in less femoral component flexion than the traditional intramedullary axis

During robotic and computer-navigated primary total knee arthroplasty (TKA), the center of the femoral head is utilized as the proximal reference point for femoral component position rather than the intramedullary axis. We sought to analyze the effect on femoral component flexion–extension position between these two reference points. We obtained CT 3D-reconstructions of 50 cadaveric intact femurs. We defined the navigation axis as the line from center of the femoral head to center of the knee (lowest point of the trochlear groove) and the intramedullary axis as the line from center of the knee to center of the canal at the isthmus. Differences between these axes in the sagittal plane were measured. Degree of femoral bow and femoral neck anteversion were correlated with the differences between the two femoral axes. On average, the navigated axis was 1.4° (range, −1.4° to 4.1°) posterior to the intramedullary axis. As such, the femoral component would have on average 1.4° less flexion compared with techniques referencing the intramedullary canal. A more anterior intramedullary compared with navigated axis (i.e., less femoral flexion) was associated with more femoral bow (R2 = 0.7, P