Abstract
Introduction
Coronal misalignment of the lower limbs is closely related to the onset and progression of osteoarthritis. In cases of severe genu varus or valgus, evaluating this alignment can assist in choosing specific surgical strategies. Furthermore, restoring satisfactory alignment after total knee replacement promotes longevity of the implant and better functional results. Knee coronal alignment is typically evaluated with the Hip-Knee-Ankle (HKA) angle. It is generally measured on standing AP long-leg radiographs (LLR). However, patient positioning influences the accuracy of this 2D measurement. A new 3D method to measure coronal lower limb alignment using low-dose EOS images has recently been developed and validated. The goal of this study was to evaluate the relevance of this technique when determining knee coronal alignment in a referral population, and more specifically to evaluate how the HKA angle measured with this 3D method differs from conventional 2D methods.
Materials and methods
70 patients (140 lower extremities) were studied for 2D and 3D lower limb alignment measurements. Each patient received AP monoplane and biplane acquisition of their entire lower extremities on the EOS system according the classical protocols for LLR. For each patient, the HKA angle was measured on this AP X-ray with a 2D viewer. The biplane acquisition was used to perform stereoradiographic 3D modeling. Valgus angulation was considered positive, varus angulation negative. Student's T-test was used to determine if there was a bias in the HKA angle measurement between these two methods and to assess the effect of flexion/hyperextension, femoral rotation and tibial rotation on the 2D measurements. One operator did measurements 2 times.
Results
The average total dose for both acquisitions was 0.75mGy (± 0.11mGy). The 2D and 3D measurements are reported in table 1. Intraoperator reliability was >0,99 for all measurements.
In the whole series, 2D–3D HKA differences were >2° in 34% of cases, >3° in 22% of cases, >5° in 9% of cases and >10° in 3% of cases >10°.
We compared 2D and 3D measurements according to the degree of flessum/recurvatum (> or <5° and > or <10°). The results are reported in table 2. The statistical analysis of parameters influencing 2D/3D measurements is reported in table 3.
Discussion and conclusion
The HKA angle is typically assessed from 2D long-leg radiographs. However, several studies highlighted that 2D assessment of this angle may be affected by patient's positioning. Radtke showed that lower limb rotation during imaging significantly affected measurements of coronal plane knee alignment. Brouwer showed that axial rotation had an even greater effect on the apparent limb alignment on AP radiographs when the knee was flexed. This last finding is particularly relevant as many lower extremities present some amount of flexion or hyperextension, especially in aging subjects. This low dose biplanar EOS acquisitions provide a more accurate evaluation of coronal alignment compared to 2D, eliminating bias due to wrong knee positioning. This study points out the interest of EOS in outliers patients and opens new perspectives for preoperative planning and postoperative control of deformity correction or knee joint replacement.
