Study population
For studying the incidence of FT after cementless THA, 40 postoperative computed tomography (CT) scans were analyzed by a single investigator (LP). This study was conducted after authorization by the Institutional Ethical Board (No. 06/100) and the Federal Office for Radiation Protection (Z5-22462/2-2007-008) and informed consent was obtained from all patients. Average patient age was 69 (±4.8) years and average body mass index (BMI) 26.4 (±3.7) kg/m2. Exclusion criteria were arthritis secondary to hip dysplasia, post-traumatic deformities of the pelvis, and - because a post-operative pelvic CT scan was required - age below 50 years at the time of surgery. All operations were done with the patient in lateral position through a modified Smith-Petersen (Micro-Hip®) approach [11] by two surgeons (TR, ES). Press-fit components and cement-free hydroxyapatite-coated stems (Pinnacle cup, Corail stem, DePuy, Warsaw, IN, USA) were used in all cases. The CT data sets (pelvis and femoral condyles) included 16 male and 24 female, 17 left and 23 right hips. None of the patients had a THA on both sides.
Alignment of stem implant and planning of landmarks
A three-dimensional (3D) CT analysis software (Hip CT 3.5.2, Brainlab AG, Feldkirchen, Germany) was used to evaluate the incidence of FT. The software included an implant database with computer-aided design (CAD) 3D models of the implant geometries provided by the implant manufacturer. The models of the actual implants were superimposed onto the image data to determine their exact position, i.e. the implants were manually aligned until the geometric models fitted optimally to the actual implants (Figure 2). According to the implant data stored in the database, the orientation of the neck and shaft axis of the stem was then assessed and analyzed. In situ orientation was determined by anatomical landmarks. The center of the implant’s head was determined by manually placing a sphere around the head in the image data. The transepicondylar axis was constructed between the most prominent aspects of the femoral condyles, visible in the horizontal plane. The mechanical axis of the femur was determined by the connecting line between the center of the femoral head and the midpoint of the transepicondylar axis. The posterior condyles were planned as the most posterior points of the femoral condyles in the CT data sets. This axis was used to complete the coordinate system of the femur, i.e. the coronal plane was defined as the plane spanned by the mechanical and the direction of the posterior condyle axis. Femoral Tilt (FT) was then calculated as the deviation between the mechanical axis and shaft axis of the stem in a sagittal projection.
Creation of a geometric model to represent the orientation of the stem implant
For the second part of the analysis, a three-dimensional (3D) computer model of the hip was used to systematically analyze changes in the femoral anatomy and its effects on femoral antetorsion (AT). In our model, we defined FT as the angle, which directly reflects the deviation between the proximal femoral shaft and the mechanical axis in a sagittal projection. Figure 3 shows the construction of the model including an initial FT (iFT) and initial AT (iAT) reflecting angles derived from a rotational representation of the neck axis alignment. According to Yoshioka [12], AT was defined as the deviation between the neck axis and the posterior condyle axis when projected to a plane orthogonal to the mechanical axis. As a first step, the effect of changes in iFT on the resulting AT angle was analyzed (Additional file 1). For this purpose, iFT values were stepwise increased from 2.1° to 9.3° for three values of iAT (0°, 15°, 30°). Following our analysis of FT on the postoperative CT scans, an iFT angle of 5.7° was considered to be an average value, whereas 2.1° and 9.3° represented lower and upper margins. VV was fixed to 4.5° in these experiments. Further on, differences between the initial (iFT) and true FT values were analyzed in the range 2.1° to 9.3° iFT and 0° to 10° VV. Respectively the zone of impingement-free, compliant cup positions was determined for stem positions with varying iFT according to the approach as published by Widmer [8]. This so called “zone-of-compliance” contains a combined, stem/cup orientation to position both components in a way that the normal range-of-motion (ROM) is contained within an impingement-free prosthetic ROM, aiming for at least 130° flexion, 40° extension, 50° abduction, 50° adduction, 40° external, and 80° internal prosthetic rotation. Additionally, an impingement-free prosthetic ROM for int./ext. rotation at 90° of flexion was taken into account. The limits were set to 45° for internal and 55° for external rotation. This intended prosthetic ROM is larger than the movements in commonplace maneuvers known to increase the risk for dislocation in THA [2, 3]. As proposed by Widmer et al., a restriction of the inclination (≤ 50°) was used as an additional constraint. ROM was calculated by an algorithm which determines collisions between the femoral and cup implant. This algorithm was implemented into a prototype software (Hip Storage Viewer, Brainlab AG, Feldkirchen, Germany). The geometry of the implants was specified by 3D CAD files (geometric 3D models) provided by the implant manufacturer. Conventional cementless implants [Pinnacle cup, standard Corail stem (NSA 135°), neutral polyethylen liner, short 32 mm head; DePuy, Warsaw, IN, USA] were used. ROM was calculated based on a neutral position of the leg by aligning the femoral and the pelvis coordinate system. The analysis did not depend on a particular coordinate system of the pelvis since only the orientation of the implants was considered.
Additional file 1: Dynamic model of the Femoral Tilt. FT is continuously increased from 0° to 10°. The resulting change of antetorsion is represented by the red line. (MP4 941 KB)
Comparison between femoral tilt and zones-of-compliance
As a last step, resulting zones-of-compliance were compared for two variations in iFT (2.1°, 9.3°). The same (effective) AT of 15° was used for this analysis, i.e. the antetorsion was adapted according to the variations in iFT. This allowed assessing differences in the zones-of-compliance when equal (effective) antetorsion values but different iFT values are given. In analogy to Widmer et al., the optimum cup position was determined as the point with the lowest inclination within the zone-of-compliance where a safety zone of approximately 1° was respected. All cup orientation angles were calculated in terms of the radiographic definition according to Murray [13].
Statistical analysis
Statistical analyses were performed using Microsoft Excel (Microsoft Inc, Redmond, WA, USA). Mean values, standard deviations, ranges, and confidence intervals with 95% confidence level were calculated. Statistical differences between the group of male and female patients were analyzed by a Student’s two sample t-tests assuming equal variances (significance level: 5%).