BackgroundThe coronal whole-leg radiograph is generally used for preoperative planning in total knee arthroplasty. The distal femoral valgus angle (DFVA) is measured for distal femoral bone resection using an intramedullary guide rod. The effect of coronal and sagittal femoral shaft bowing on DFVA measurement in the presence of malrotation or knee flexion contracture has not been well reported. The objectives of this study were: (1) to investigate the effects of whole-leg malrotation and knee flexion contracture on the DFVA in detail, (2) to determine the additional effect of coronal or sagittal femoral shaft bowing.MethodsWe studied 100 consecutive varus and 100 valgus knees that underwent total or unicompartmental knee arthroplasty. Preoperative CT scans were used to create digitally reconstructed radiography (DRR) images in neutral rotation (NR, parallel to the surgical epicondylar axis), and at 5° and 10° external rotation (ER) and internal rotation (IR). The images were also reconstructed at 10° femoral flexion. The DFVA was evaluated in each DRR image, and the angular variation due to lower limb malposition was investigated.ResultsThe DFVA increased as the DRR image shifted from IR to ER, and all angles increased further from extension to 10° flexion. The DFVA variation in each position was 1.3° on average. A larger variation than 2° was seen in 12% of all. Multivariate regression analysis showed that sagittal femoral shaft bowing was independently associated with a large variation of DFVA. Receiver operating characteristic analysis showed that more than 12° of sagittal bowing caused the variation.ConclusionIf femoral sagittal bowing is more than 12°, close attention should be paid to the lower limb position when taking whole-leg radiographs. Preoperative planning with whole-leg CT data should be considered.