Variability of three-dimensional knee morphology cannot be effectively assessed using a coronal plane knee alignment...

Optimal reproduction of the native three-dimensional (3D) alignment in total knee arthroplasty (TKA) influences outcomes; however, much of the modern TKA alignment research, such as the coronal plane alignment of the knee (CPAK), focuses only on coronal alignment. Tibial, femoral and tibiofemoral measurements on the axial and sagittal planes were evaluated for their relationship to the arithmetic hip-knee-ankle angle (aHKA) and joint-line obliquity (JLO). These 3D anatomical measurements are also evaluated across CPAK groups.

A retrospective analysis of the 360 Med Care computed tomography (CT) database was performed. Patient CT scans were segmented and landmarked. Linear regression analysis compared 12 axial and sagittal plane measurements (representing tibial, femoral and tibiofemoral rotation, tibial slope and femoral flexion) with both aHKA and JLO. Nonparametric tests assessed these anatomical measurements across the different CPAK groups, while Cohen's delta (d) determined the effect size.

With a sample size of 7450 osteoarthritic knees, significant but weak relationships (r < 0.30) were observed between all 12 anatomical measurements and both aHKA and JLO. Tibiofemoral rotations between Insall's axis and both the posterior condylar and the surgical transepicondylar axes demonstrated large effect sizes (d > 0.80). However, trivial to small effect sizes (d < 0.50) were broadly observed across the 12 axial and sagittal measurements, underscoring their limited clinical significance.

While useful for describing coronal knee anatomy, CPAK phenotypes fail to differentiate tibial, femoral and tibiofemoral rotation, tibial slope or femoral flexion-crucial aspects of 3D surgical planning. Therefore, more comprehensive knee phenotyping solutions are required to guide individualised TKA alignment strategies.

Level II.