Abstract
Summary sentence
The bowing of the femur defines a curvature plane to which the proximal and distal femoral anatomic landmarks have a predictable interrelationship. This plane can be a helpful adjunct for computer navigation to define the pre-operative, non-diseased anatomy of the femur and more particularly the rotational alignment of the femoral component in total knee arthroplasty (TKA).
Background and aims
There is very limited knowledge with regards to the sagittal curvature -or bowing- of the femur. It was our aim (1) to determine the most accurate assessment technique to define the femoral bowing, (2) to define the relationships of the curvature plane relative to proximal and distal anatomic landmarks and (3) to assess the position of femoral components of a TKA relative to the femoral bowing.
Materials and methods
Four independent algorithms were developed and tested on 3D models of 18 cadaveric femora. A sensitivity study showed that a bisector-based method supplied the most stable results. In order to verify if the curvature plane can be used for TKA alignment, the anteversion angle was determined relative to this plane and compared with anteversion angles defined using the coronal plane.
Results
The average curvature of the cadaveric femora was 895.85 mm (SD = 184.53 mm).
The mean anteversion angle calculated along the projected mechanical or anatomical axis in the coronal plane were 8.2+/−5.2° and 7.6+/− 4.8°. These angles calculated along the projected mechanical or anatomical axis in the curvature plane were 8.2+/−5.2° and 5.2+/−4.8° respectively (p>0.05).
Assessment of the component placement relative to the mechanical axis showed that in the coronal plane, an average deviation of 1.84° was measured. In the sagital plane, the average deviation from the mechanical axis was 2.01°. The components were placed in 1 to 2° of extension relative to the femoral bowing.
Discussion
A new and stable algorithm was successfully developed to determine the curvature of the femoral shaft. This curvature was comparable to 2 previously reported curvatures.
Our study also demonstrates a predictable interrelationship between the femoral shaft curvature on one side and the rotation of the distal femur on the other side. This finding is of great interest in view of a recent trend amongst knee surgeons to aim at anatomical restoration of the patient's original anatomy. Patient matched cutting blocks as well as patient specific implants are today increasingly considered in daily practice in an attempt to restore the patient's natural anatomy and biomechanics. Computational methods to reverse engineer the pre-diseased status of the knee joint regarding its anatomy and orientation are therefore of great importance. The findings from our study suggest that the femoral shaft's curvature is a helpful adjunct to this. Furthermore, abnormal rotational alignment of an axially malaligned component can be assessed accurately with this new reference plane. However, further research on implementing this algorithm and this plane into clinical practice is mandatory. However, further research on implementing this into clinical practice is required.
