The length of lateral radiographs significantly impacts the measurement of the femoral intramedullary axis in patients undergoing total knee arthroplasty

Abstract

Background Accurate femoral component alignment in the sagittal plane is crucial for total knee arthroplasty (TKA) success. In manual TKA, sagittal alignment is typically guided by the intramedullary axis (IMA) determined on lateral radiographs. However, due to femoral bowing, the IMA varies along the femoral shaft, raising the question of the opti mal level for referencing this axis. As short-segmented knee radiographs (SSKR) are increasingly used in clinical practice, it is unclear whether they introduce systemic deviations in IMA determination. This study aimed to compare the IMA derived from SSKR and conventional lateral radiographs (CLR), and to assess whether axis deviation increases with femoral shaft length. Methods This retrospective analysis included 153 patients undergoing primary TKA. The femoral IMA was deter mined using a two-circle method on both the full CLR and a 12.5 cm distal segment simulating SSKR. For the CLR axis, one circle was positioned at the most proximal point of the femoral shaft visible on the radiograph, and the second circle was placed 5 cm proximal to the distal femoral joint line. For the SSKR-based axis, the distal circle remained identical, while the proximal circle was repositioned 12.5 cm proximal to the joint line. Measurements were performed twice by two observers. The angular deviation between CLR- and SSKR-based axes was reported in degrees. A one sample t-test was used to test for statistical significance. Clinically relevant deviation was defined as≥2°, and the num ber and percentage of such outlier cases were reported. Correlation between femoral shaft length and angular devia tion was analyzed using Pearson correlation. A multivariable regression tested whether femoral length independently predicted angular deviation after adjusting for age, sex, BMI, and side. Results The IMA on SSKR was significantly more posterior than on CLR, with a mean angular deviation of 2.3°±1.1° (95% CI: 2.2–2.5; P<0.001; Cohen’s d=2.1). In 57 cases (38%), deviation exceeded the clinically relevant threshold of≥2°. A significant positive correlation was found between the visible femoral shaft length and the angular deviation between CLR and SSKR axes (r=0.504, P<0.001). In multivariable regression, femoral length remained an independent predictor of angular deviation after adjustment for age, sex, BMI, and side (P<0.001). Conclusion Referencing the IMA on SSKR results in a significantly more posterior axis compared to CLR, which may lead to increased femoral component flexion in TKA. Given the high incidence of outlier cases and their association with femoral shaft length, surgeons should be cautious when relying on short radiographs for preoperative planning of sagittal femoral alignment.