Purpose of the study: Radiographs of 24 patients who underwent surgery for total hip arthroplasty (THA) with a locked stem were reviewed at 38 months mean follow-up using a dedicated software. This software enables digital analysis of standard radiographs with semiquantitative evaluation of bone density.

Material and method: Good quality postoperative AP views of the femur and the same view at last follow-up were selected using the same criteria. These images were digitalised then analyses with the software. Bone density was established along a horizontal line 1 cm below the lesser trochanter perpendicular to the femur shaft. Computer analysis of bone density established three categories of patients as a function of cortical density: no cortical modification (n=5 hips), modification of only one cortical (n=11) and modification of both corticals (n=8).

Results: Bone density increased, suggesting improve cortical bone stock as has been reported by most authors using the transfemoral approach and a non-cemented locked stem.

Discussion: This result confirms the data in the literature; data which, unlike our series, were established on qualitative or subjective evaluations. The method presented here has the advantage of a semi-quantitative analysis, simple use, applicable to plain x-rays, and good reproducibility since all measures are made by the software. This study demonstrated the notion of cortical quality since it was not limited to a simple measurement of width, but also bone density, closer to real intraoperative observations.

Conclusion: Use of this method enables longitudinal study to establish the kinetics of bone remodelling, compare results between surgical methods, and search for factors explaining observed variations.

Purpose of the study: Recent studies have demonstrated that navigation systems provide highly accurate cuts for orthogonal alignment of the lower limb. The accuracy has not to our knowledge been assessed for rotation. Rotation of the femoral piece, which results from a strategy independent of the bone cut, is designed to «correct» for epiphyseal torsion of the distal femur and thus obtain a biepicondylar axis parallel to the «surgical» posterior bicondylar line described by Berger (line drawn between the medial sulcus and the lateral epicondyle), i.e. forming un angle of 2° with the anatomic biepicondylar line described by Yoshioka (line from the medial to lateral condyles). The purpose of this study was to access the precision of navigation rotation.

Material and methods: This prospective consecutive study included 40 osteoarthritic knees undergoing total knee arthroplasty (TKA). The anatomic angle of distal femoral torsion (Yoshioka angle: angle formed by the posterior bicondylar line and the biepicondylar line) was measured on the pre- and post(3 months)-operative scans. Navigation (Navitrack, Zimmer) used the rotation given by the preoperative scan to guide the femoral cut with the objective of achieving a residual Yoshioka angle of 2°, i.e. parallel to Berger’s surgical biepicondylar line. The postoperative HKA measured on the pangonogram in the standing position was 179.6±2° with 85% of patients between −2° and +2°, confirming the reliability of the navigation system.

Results: The mean preoperative epiphyseal rotation of the distal femur was 6.4±1.8°. The mean postoperative measurement was 1.1±2.4°. Eighty percent of patients were within ±2° of the objective.

Discussion: We demonstrated in previous work that navigation-based rotation using intraoperative data is satisfactory as long as the degree of rotation is based on the preoperative scan (and thus takes into account the wide rang of distal femur torsion). Navigation-based rotation is a progress compared with standardized rotation. The few errors observed were related to insufficient identification of the posterior bicondylar line during navigation or to difficulties in interpreting the postoperative scan.

Purpose of the study: The purpose of rotating the femoral piece, using an indepenent cut strategy, is to «correct» for epiphyseal torsion of the distal femur and thus obtain a biepicondylar axis parallel to the posterior bicondylar axis. It is known however that epiphyseal torsion of the distal femur is highly variable from one individual to another. Intraoperative identification of the biepicondylar line enables appropriate rotation, as long as the data collected are reliable. The purpose of this study was to determine the reliability of intraoperative biepicondylar axis measurements made with navigation systems and to compare the results with the preoperative scan taken as the gold standard.

Material and methods: This prospective study included 60 degenerative knees undergoing total knee arthroplasty. The angle of epiphyesael rotation of the distal femur was measured on the preoperative computed tomography scan and intraoperatively with the navigation system which identified the biepicondylar line and the posterior bicondylar line. Statistical regression lines were determined.

Results: The rotation measured on the preoperative scan was 7.1±2.4° and by the intraoperative navigation system 3.2±4.3°. There was a very weak statistical correlation between the preoperative measurement and the intraoperative navigation measurement (p=0.234, R =0.320).

Discussion: Intraoperative identification of the biepicondylar axis is not reliable. Navigation does not enable an accurate assessment of the distal epiphyseal torsion of the femur and thus the proper rotation to give to the femoral piece. The only reliable measurement of the epiphyseal rotation of the distal femur is made on the preoperative computed tomography.