Analysis of factors that affect the precision of the radiographic lateral femoral bowing angle using a three-dimensional computed tomography-based modelling technique

Precise measurement of lateral femoral bowing is important to achieve postoperative lower limb alignment. We aimed to investigate factors that affect the precision of the radiographic lateral femoral bowing (RLFB) angle using three-dimensional (3D) models and whether the angle affects surgery design. Forty femurs in total were divided into two groups based on their preoperative RLFB angle. The flexion contracture angle, preoperative and postoperative RLFB angles, and intersection angle between the mechanical and anatomical axes were compared. The angle between the arc and sagittal planes, varus and valgus angles, and intersection angle between the mechanical and anatomical axes were measured on a 3D model. There was no significant between-group difference in 3D model measurements of the angle between the arc and sagittal planes (p = 0.327). There was no significant difference between the mechanical and anatomical axes measured by both imaging modalities (p > 0.258). When the RLFB was >5°, the flexion contracture angle and radiographic femoral bowing angle were positively correlated (r = 0.535, p < 0.05). Distal femur varus and valgus angles significantly differed between the two groups (p = 0.01). After total knee arthroplasty, the radiographic femoral bowing angle decreased significantly. When the cases’ radiographic femoral bowing angle is larger and the angle between the arc and sagittal planes is smaller as measured in 3D models, the angle between the arc and coronal planes is larger. The radiographic femoral bowing angle does not reflect the actual size of lateral femoral bowing, does not greatly affect surgery design, and is greatly affected by flexion contracture deformity. A RLFB angle larger than 15° indicates real lateral femoral bowing.