Effects of rotation on measurement of lower limb alignment for knee osteotomy.

The purposes of this study were to clarify the effects of rotation on two-dimensional measurement of lower limb alignment for knee osteotomy using a three-dimensional method and to determine whether this 3-D simulation method could help with planning of knee osteotomy. We developed computer software to calculate femorotibial angle (FTA) and hip-knee-ankle angle (HKA) and simulate knee osteotomy from a CT-based 3-D bone model of the lower limb. Lower limb rotation on anteroposterior long-standing radiographs was measured by superimposing the 3-D bone models. Changes in alignment with limb rotation were calculated using the software. FTA after virtual closed-wedged osteotomy was measured for a hypothetical case of a rotation error of the osteotomy plane in reattaching the proximal cutting surface to the distal cutting surface. For 31 varus knees in 20 patients with medial compartment arthritis, the mean rotation angle, relative to the epicondylar axis, with variable limb position was 7.4 +/- 3.9 degrees of internal rotation (mean +/- SD), ranging from 8 degrees of external rotation to 14 degrees of internal rotation; the mean changes in FTA and HKA were 3.5 +/- 2.2 degrees (range, 0.4-8.6) and 1.6 +/- 1.3 degrees (range, 0.2-4.9), respectively. The FTA "flexion angle" (lateral view alignment from neutral AP) and the absolute HKA "flexion angle" correlated with the change in FTA and HKA with limb rotation, respectively (FTA, R = 0.999; HKA, R = 0.993). The mean change in FTA after virtual closed-wedged osteotomy was 3.2 degrees for internal and external 10 degrees rotation errors in reattaching the osteotomy plane. Rotation may affect measurement of lower limb alignment for knee osteotomy, and 3-D methods are preferable for surgical planning.