[Distal femoral torsion: comparison of CT scan and intra operative navigation measurements during total knee arthroplasty. A report of 70 cases].

INTRODUCTION: Proper positioning of the prosthesis components in total knee arthroplasty is an important factor for satisfactory outcomes such as mobility, pain and wear. If an independent cutting technique is performed, rotational alignment of the femoral component should adapt patient's specific anatomy. The distal epiphyseal femoral torsion (DEFT) is adjusted in order to align the prosthetic posterior condylar axis along the bone transepicondylar axis. The DEFT presents a high rate of interindividual variations. Computed tomography scanning produces reliable and reproducible measurement of the epiphyseal torsion, but this requires additional procedure. We therefore used intraoperative computed navigation for DEFT measurement, thus accurately adapting each patient's epiphyseal torsion during the procedure, without resorting to the preoperative CT scan.

MATERIAL AND METHODS: This prospective study included 70 patients with arthritic knees who underwent TKA. Mean patient age was 74 years old. There were 52 women, 35 right knees, 33 genu varum, 19 genu valgum, 18 normal knees. DEFT was determined by preoperative CT scan, using the Yoshioka angle referencing. Distal epiphyseal femoral torsion measurement was carried out using the Navitrack system (Orthosoft). The DEFT was the navigated measured angle between the transepicondylar axis and the posterior condylar axis manually located and digitized with an optically tracked stylus. The navigation system was therefore used as a simple digital measurement device to evaluate the distal epiphyseal femoral torsion. We had established the reference transepicondylar axis as the line connecting the prominence of the medial and lateral epicondyles. During the procedure, we also carried out computer-assisted measurement of HKA axis in full extension and at 90 degrees of knee flexion. Correlation between the navigated HKA in full extension and HKA measured on the preoperative pangonogram (R(2)=0.621) demonstrated a high reliability of the navigation system in the frontal plane.

RESULTS: There was no correlation between the mean epiphyseal torsion determined with computer navigation and the epiphyseal torsion measured on the CT scan (R(2)=0.09). Significant interindividual variations were reported. Navigated HKA at 90 degrees of knee flexion was not correlated with navigated HKA in full extension (R(2)=0.398) nor with epiphyseal torsion measured on the CT scan (R(2)=0.063). Results demonstrated a major interindividual variation.

DISCUSSION: Our results report a large variability in distal epiphyseal femoral torsion measured with CT scan. Moreover, computed navigation does not provide a reliable and reproducible evaluation of the epiphyseal torsion. Due to inaccurate identification of femoral epicondyles, the related navigated measurement is not considered to be reliable and reproducible. Navigated HKA at 90 degrees of knee flexion is not a fair indirect reflection of epiphyseal torsion. Computer-assisted navigation fails to provide direct or indirect, reliable and reproducible intraoperative measurement of distal epiphyseal femoral torsion. Preoperative CT scan is the only reliable method to produce accurate measurement of distal epiphyseal femoral torsion.