Do we need hip-ankle radiographs to assess the coronal alignment and implant position after total knee replacement?

AIMS AND OBJECTIVES: Restoration of the coronal alignment of the knee is known to be one of the major criteria of a successful total knee arthroplasty (TKA). It therefore appears to be mandatory to routinely assess the postoperative limb alignment using hip-ankle radiographs and to identify implants that may be at risk of premature failure. However, there is no clear consensus whether weight-bearing hip-ankle radiographs or rather standardized a-p knee-radiographs should be used to assess implant position and coronal alignment after TKA. It is the aim of the present study to investigate if implant position and the mechanical alignment after TKA can reproducibly be assessed using standardized a-p knee-radiographs or rather if weight-bearing hip-ankle radiographs are needed. MATERIALS AND METHODS: This study was performed on 100 postoperative weight-bearing hip-ankle radiographs after conventional primary TKA. The true mechanical and anatomical femorotibial angle as well as coronal implant position (MPTA, LDFA) was assessed using the MediCAD software, which served as a control. The hip-ankle radiographs were then digitally cropped to 80%, 60% and 40% of the leg-length. In each cropped radiograph, tibial coronal implant position was assessed by referencing against the visible mid-shaft, whereas femoral implant position was referenced against the visible mid-shaft (anatomical axis) or against a surrogate mechanical axis, which was drawn perpendicular to the distal tangent of the femoral component. Each measurement was performed by three independent observers. The difference between the alignment parameters in the hip-ankle radiographs were statistically compared with the cropped radiographs and the inter-observer correlation coefficient (ICC) was calculated for each parameter. RESULTS: The ICC for inter-observer agreement of measurement of the mechanical femorotibial angle was significantly higher in hip-ankle radiographs (.95) when compared with a radiograph cropped to 80% (.89), 60% (.76) or 40% (.61). The difference in the mechanical femorotibial angle between a hip-ankle radiograph and a radiograph cropped to 40% averaged 2.2° (+/- 1.6°), this difference being significant (p<0.001). There were no significant differences between hip-ankle radiographs and any cropped radiographs when comparing coronal implant position and femorotibial angle as referenced against the anatomical axes. CONCLUSION: We conclude that compared to hip-ankle radiographs, measurements from standard a-p knee-radiographs are not sufficient to accurately and reproducibly assess the mechanical alignment following TKA. However, standard a-p knee-radiographs are appropriate to assess implant position when referenced against the anatomical axes. Since the literature provides compelling evidence that implant survival is predicted by the varus-valgus position of the femoral and tibial component rather than by the overall mechanical axis, it should be discussed if weight-bearing hip-ankle radiographs should be abandoned as a routine after TKA.