Discrepancies in long-leg alignment and knee joint line obliquity between two- and three-dimensional measurements under weight-bearing conditions: Effects on coronal plane alignment of the knee classification

Background: Long-leg alignment and joint line obliquity have traditionally been assessed using two-dimensional (2D) radiography, but the accuracy of this measurement has remained unclear. This study aimed to evaluate the accuracy of 2D measurements of lateral distal femoral angle (LDFA) and medial proximal tibial angle (MPTA) using upright three-dimensional (3D) computed tomography (CT). Methods: This study involved 66 knees from 38 patients (34 women, four men) with knee osteoarthritis (OA), categorized by Kellgren–Lawrence (KL) grade. Patients underwent standing long-leg radiography (LLR) and upright CT. Flexion and rotation angles of the femur and tibia were measured from upright CT data, and knees were classified by coronal plane alignment of the knee (CPAK) according to 2D- and 3D-LDFA/MPTA. Results: Overall, as KL grade increased, femoral external rotation and flexion increased significantly (P < 0.05). Regarding the tibia, flexion increased (P < 0.05) while rotation remained unchanged with increasing KL grade. In mild OA (KL1 or KL2), 2D-LDFA (86.5 ± 1.8°) was significantly larger than 3D-LDFA (85.0 ± 2.5°; P < 0.05), and this difference was also observed in severe OA (KL3 or KL4) (88.7 ± 2.5° vs. 87.7 ± 3.2°; P < 0.05). However, MPTA was comparable between 2D and 3D. The consistency between 2D and 3D CPAK classifications was 48.5% for the entire sample, 25.0% for mild OA, and 61.9% for severe OA. Conclusion: Based on 3D-LDFA/MPTA data gained from upright CT, 2D radiographic LDFA tended to be inaccurate, which may significantly affect the CPAK classification.