Although proximal tibia vara is physiologically and pathologically observed, it is difficult to measure the varus angle accurately and reproducibly due to inaccuracy of the radiograph because of rotational and/or torsional deformities. Since tibial coronal alignment in TKA gives influence on implant longevity, intra- or extra-medurally cutting guide should be set carefully especially in cases with severe tibia vara. In this context, we measured the proximal tibial varus angle by introducing 3D-coordinate system.

Bi-cruciate substituting total knee arthroplasty (TKA) having two post-cam mechanisms was developed to substitute for cruciate ligament function after surgery. A previous study has shown many of these knees achieve high functional flexion. However, there is little information provided to differentiate between knees able to flex deeply and those that could not, although this is a major concern for surgeons. This study was conducted to compare the kinematic pathway from 0° to 90° in both groups.

Twenty five knees were included in this study. All knees were diagnosed with osteoarthritis (OA) and all TKAs were performed by the same surgeon (WR) from November 2005 to September 2006. A mini mid-vastus surgical approach with posterior cruciate ligament (PCL) resection and patellar resurfacing was used in all cases. Computer navigation was used to guide bone cuts in all the cases. Patients' age averaged 63 years (range, 43–73) at the time of surgery. The study observations were performed at an average of 53 (SD 4) months after surgery. Knee motions were recorded using video-fluoroscopy while subjects performed stair up and down, and lunge activities. The three-dimensional position and orientation of the implant components were determined using model-based shape-matching techniques. This initial manual solution was refined using nonlinear least-squares optimization to maximize image-edge correspondence. Joint kinematics were determined from the three-dimensional pose of each implant component using Cardan/Euler angles. TKAs were divided into two groups according to the maximum lunge angles; TKAs achieved larger than 130° were defined as high flexion group (H group) and the ones from 110° to 130° were defined as moderate flexion group (M group). Tibial internal position and the AP locations of medial and lateral condyles were examined.

Two TKAs were excluded since their maximum flexion was less than 110°. Twelve and eleven TKAs were defined as the H group (High flexing, average 137°, SD 4°) and the M group (Moderate flexing, average 121°, SD 5°), respectively. Tibial internal rotation averaged 10° (SD 4°) and 9° (SD 3°), respectively, at lunge position. The medial and the lateral condyles were located at 9 mm (SD 2 mm) and 17 mm (SD 3 mm) posterior to the tibial centerline during the lunge activity in the M group and at 11 mm (SD 2 mm) and 21 mm (SD 3 mm) in the H group. Tibial rotation was not statistically different (Figure 1), while AP position of the lateral condyle translated more backward in H group at 90° (Figure 2). The TKAs in the M group exhibited femoral forward motion from 0° to 20° flexion, while the H group moved backward (Figure 2).

Our results revealed the post-cam mechanisms worked effectively in the H group TKA. The TKAs which acquired deep flexion successfully prevented the “roll forward motion” and had greater femoral posterior translation at 90° where the posterior post-cam mechanism engages. It appears adequate femoral posterior translation may be important to acquire deep flexion after TKA.

Although optimal alignment is essential for improved function and implant longevity after TKA, we have less bony landmarks of tibia relative to femur. Trans-malleolar axis (TMA) is a reference line of distal tibia in the axial plane, which externally rotated relative to a ML axis of proximal tibia. We originally defined another reference axis associated with the orientation of tibial plafond, and then measured tibial torsion in the 3D-coordinate system.

Three-dimensional CAD models of 20 tibiae were reconstructed based on pre-operative CT data from OA patients (16 females and 4 males, 73.8 ± 6.9 years old). TMA was a line connecting each apex of medial and lateral malleolus. The plafond axis (PLA) that we originally defined in this study was a line connecting each midpoint of medial and lateral margin of talocrural facet. In terms of interobserver correlation coefficiency and mean errors of the designated points to define those axes, TMA was found out to be 0.982, 3.14 ± 0.47 mm (medial), and 0.988, 4.88 ± 0.59 mm (lateral). Those of PLA were 0.997, 1.97 ± 0.53 mm (medial), and 0.995, 2.02 ± 0.44 mm (lateral). The tibial torsion was 16.3 ± 6.3°with reference to TMA, and 10.2 ± 8.4°to PLA.

Based on these results, as for the rotational reference axis in the axial plain of distal tibia, we consider the plafond axis to be another reliable and reproducible axis, which is expected to be applicable in preoperative planning in TKA to reduce outliers of coronal alignment.