MEASUREMENT OF THREE-DIMENSIONAL TOTAL KNEE ARTHROPLASTY ALIGNMENT WITH COMPUTERISED TOMOGRAPHY

Background The use of Computed Tomography (CT) as a medical imaging tool has widespread applications in the field of knee surgery. Surgeons use a CT scan in a conventional way during the pre-operative stage, to plan the position of the femoral component in the horizontal plane. In the post-operative stage, the use of a CT scan is a routine tool in the evaluation of failed TKA as rotational malalignment of the femoral component has been determined as a cause of poor clinical outcome after TKA. Aim How accurately can we measure the different angles with importance for alignment on a 3D-image in comparison to a standard CT, 2D, image. Material and methods This study includes patients above 55 years of age who were scheduled for a TKA at our centre and who had a pre- and postoperative full-leg length computed tomography (CT). These images were analysed using Mimics V 16.0 ® and 3-matic V 8.0 ® (Materialise, Haasrode, Belgium) to create the surface reconstruction and perform the 3D-measurements. Different angles were measured pre- and post-operatively on these images both in 2D as in 3D: condylar twist angle (CTA), posterior condylar angle (PCA), hip-knee-ankle angle (HKA), tibiofemoral rotation angle (TFRA), posterior tilt of the tibial implant (STPA) and the frontal plane angle of the tibial implant (FTPA). A power analysis showed a needed sample size of 18 patients. Pre-operatively 21 patients were included, 18 of them also received a post-operative full-leg CT. Three observers participated in the study and they all performed all analyses twice with a minimum interval of one week for obtaining intra-observer repeatability. Statistical analysis was performed to obtain the intra- and interobserver variability. Results The intra-and interobserver intra-class coefficients (ICC) were evaluated using the classification of LandisK ICC PCA=0,66. Rotation in 2D: ICC PCA=0,75, ICC CTA=0,84. The highest ICC was found for the HKA: ICC>0,98 both 2D and 3D. Postoperative results All ICC9s are in the highest category (‘almost perfect’) except for the intraobserver ICC PCA in 2D which was ‘substantial’. Conclusions and clinical implications CT-evaluation is invaluable for the preoperative planning of the position of the femoral component, for the evaluation of the rotation of the femoral component and for accurate assessment of the overall knee alignment,. This study demonstrates low intra- and inter- observer variability in the CT measurement of the pre- and postoperative alignment of the knee. To determine the rotation of the femoral component preoperatively there was no advantage for 3D over 2D. Post-operatively the use of the 3D technique has a lower variability with regard to the assessment of the rotation of the femoral component.