It is essential to understand rotational alignment of the distal femur when performing total knee arthroplasty (TKA). Several rotational landmarks including condylar twist angle (CTA) are used for preoperative planning and during TKA. Axial radiography of the distal femur is used for measuring the CTA, and assessing rotational alignment in TKA. The aim of this study was to investigate the reliability and the reproducibility of the CTA using two different methods and evaluate if CTA differed between varus and valgus knees and between normal and osteoarthritic knees.

CTA were obtained from 144 knees (77 patients) having total knee or hip arthroplasty using computed tomography (CT) and axial radiography. Subjects were divided into five groups based on femorotibial angle (FTA) and into four groups based on the severity of knee osteoarthritis. The intra-observer and inter-observer reliabilities of these methods and inter-method differences were evaluated.

The mean CTA was 7.02° with axial radiography, and 6.87° with CT images. There were no significant differences among the five FTA groups and among the four osteoarthritis groups. In total, intra-/inter-observer, and inter-method intraclass correlation coefficients were substantial or almost perfect in the scoring system of Landis et al. However, discrepancies ≥2° between the two methods were observed in more than 20% of knees.

The CTA should be reassessed by more than two observers or two methods for precise preoperative TKA planning in cases where it is difficult to identify the bony landmarks for CTA measurements.

Level III.