Native femoral flexion can be reliably measured on standard lateral knee radiographs: a reliability study with CT-based comparison - 24/09/26

Doi : 10.1016/j.otsr.2026.104874

Giancarlo Giurazza ^a, Antoine Perrier ^b, Victoria Tessier ^b, Wilfrid Graff ^b, Simon Marmor ^b, Antoine Mouton ^b, Thomas Aubert ^b,⁎

^a Orthopedics and Traumatology, Fondazione Policlinico Universitario Campus Bio-Medico, Rome, Italy

^b Groupe Hospitalier Diaconesses Croix Saint Simon, 125 Rue d’Avron, 75020 Paris, France

^⁎ Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Thursday 24 September 2026

Abstract

Background

Personalized alignment strategies in total knee arthroplasty (TKA) increasingly emphasize restoration of native anatomy in both the coronal and, more recently, the sagittal plane. However, unlike the tibial side, where preoperative and postoperative posterior tibial slope (PTS) can be directly compared using standard lateral radiographs, femoral sagittal alignment has traditionally been assessed only postoperatively as the "gamma-angle". To date, no widely adopted method exists to define native preoperative femoral flexion, preventing meaningful pre-/postoperative comparison and individualized sagittal planning.

Hypothesis

It was hypothesized that native femoral flexion could be measured on standard lateral knee radiographs using a simple and reproducible method, showing strong agreement with corresponding CT-based measurements.

Methods

This retrospective study included 44 patients (48 knees) with end-stage knee osteoarthritis. The native femoral flexion angle was measured on true lateral radiographs as the angle between the distal femoral diaphyseal axis (line connecting the midpoints of shaft levels located two and three times the femoral diaphyseal diameter from the intercondylar notch) and the meta-epiphyseal axis (line connecting the distal diaphyseal point to the center of a best-fitting circle constrained by the sulcus and Blumensaat lines). A custom automated 3D CT-based analysis was performed to obtain corresponding femoral flexion angles. Intra- and inter-observer reliability and agreement between methods were evaluated using intraclass correlation coefficients (ICCs) and Bland–Altman analysis.

Results

The mean radiographic femoral flexion angle was 2.5 ± 1.84 ° (range −2.1 ° to 6.9 °). Intra- and inter-observer reliability were excellent (ICC = 0.94 and 0.92, respectively). Agreement between radiographic and CT-based measurements was very strong (ICC = 0.86; 95% CI 0.77–0.92). Bland–Altman analysis showed a mean bias of 0.35 ± 0.9 °, with 95.9% of cases showing <2° difference between methods.

Conclusion

Native femoral flexion can be reliably measured on standard true lateral knee radiographs using a simple and reproducible method, showing strong agreement with a CT-based reference. This represents a step toward integrating sagittal plane considerations into individualized TKA planning.