The rotational position of the femoral component is a primary driver of success in total knee arthroplasty. However, distal femoral torsion (DFT) varies greatly between individuals. Measuring DFT preoperatively by CT in combination with computer-assisted surgery can significantly improve the rotational positioning of the femoral component. However, a preoperative CT scan is costly and exposes the patient to radiation. These are doubled when the patient is undergoing bilateral arthroplasty. The aim of this study was to determine the DFT in both knees of a patient undergoing bilateral arthroplasty. We hypothesized that DFT was symmetric between a patient's two knees and was independent of frontal alignment.

In this retrospective study of TKA cases performed between December 2008 and March 2015, 82 patients (mean age 73years) who underwent two-stage bilateral TKA (164 knees) were included. A preoperative CT scan of each knee was performed to measure the DFT using the surgical posterior condylar angle (PCA) described by Yoshioka. Two observers performed the measurements twice each, to allow calculation of the intraclass and interclass correlation coefficients.

The mean PCA was 5.4° (±1.48) in the right knee and 5.4° (±1.45) in the left knee, with a left/right difference ranging from 0 to 2.2° (p=0.8). In the entire cohort, 84.6% of patients had a left/right difference of less than 1°. We found no significant differences in DFT in knees with large or small frontal deformity (deformity<10°, p=0.7; deformity>10°, p=0.5) or the presence of varus or valgus (p=0.9). The intraclass correlation coefficient was excellent (94%) and the interclass correlation coefficient was moderate to good (60% for left knees, 53% for right knees).

Based on CT scan measurements, the DFT in both knees of an arthritic patient is comparable and this measurement is reproducible. This means that a single, unilateral preoperative CT scan is sufficient for planning purposes.

IV (retrospective cohort study).