Relative movements of the bones
The pattern of movement of the unloaded knee is highly ordered. In a study in which the movements were controlled solely by the intact ligaments and articular surfaces, 12 unloaded cadaver knee specimens were examined with the muscle tendons removed, the proximal tibia fixed with its plateau approximately horizontal and an intramedullary rod in the distal femur lying on a horizontal rod which was gently lowered and raised to control flexion and extension (Fig. 3.1) (Feikes 1999; Wilson 1995; Wilson et al. 2000). The only load present was the weight of the distal femur (about 5 N, 1 lb), shared between the horizontal rod and the knee specimen. An electromagnetic digitiser (Isotrack II, Polhemus Inc, Vermont, USA) was used to track all six degrees of freedom of movement of the femur relative to the tibia. The entire rig was made of plastic so as not to distort the magnetic field of the digitiser.
Figure 3.1 Fixed tibia rig used to study passive motion of the knee.
Figure 3.2 shows axial rotation of the femur relative to the tibia plotted against flexion angle for one specimen. External rotation of about 20° accompanied 120° of flexion but it is notable that the path followed during flexing was almost exactly reproduced during extending, with very little hysterisis. Axial rotation was uniquely coupled to flexion angle.
Figure 3.2 Axial rotation plotted against flexion angle for one specimen.
Figure 3.3 Axial rotation, abduction/adduction, and the three components of translation of a single point in the femur (the most proximal anterior point on the PCL attachment area) plotted against flexion angle for a single specimen (dotted curves). The solid lines on the graphs will be discussed later. (Adapted from Wilson DR, Feikes JD, Zavatsky AB, O’Connor JJ. The components of passive knee movement are coupled to flexion angle. J Biomech 2000; 33:465-73, with permission.)
The graph of axial rotation from Figure 3.2 is included again in Figure 3.3 together with plots against flexion angle of the angle of abduction/adduction and the three components of translation of a single point on the femur relative to the tibia (dotted lines). Movements are plotted relative to a coordinate system with flexion/extension calculated about a medio-lateral axis parallel to a line joining the centres of curvature of the posterior femoral condyles, axial rotation about an axis perpendicular to the medio-lateral axis and fixed in the tibia according to a method described by Yoshioka et al. (1989), ab/adduction was calculated about a ‘floating’ antero-posterior axis perpendicular to both. The three components of translation were also calculated relative to these axes.
The figures show that axial rotation, abduction/adduction and the three components of translation of an arbitrarily chosen point on the femur (in this case, the most proximal point on the attachment of the PCL) were all uniquely coupled to flexion angle. In particular, for all five degrees of freedom, the path followed during flexion was almost exactly retraced during extension, with virtually no hysteresis. These curves are therefore characteristic of the unique path of passive motion followed by this specimen. Specifying the flexion angle completely determined the configuration of this knee joint, which therefore behaved like a single degree of freedom system. The solid lines in these figures will be discussed below.
Figure 3.4 Axial rotation, abduction/adduction, and the three components of translation of a point in the femur (the most anterior point on the PCL attachment area) plotted against flexion angle for twelve specimens, the curves give mean values, the shaded areas defined by ± one standard deviation. (Reprinted from Wilson DR, Feikes JD, Zavatsky AB, O’Connor JJ. The components of passive knee movement are coupled to flexion angle. J Biomech 2000; 33: 465-73, Fig. 3, with permission.)
Figure 3.4 shows similar curves for 12 specimens tested in the same rig. All specimens exhibited minimal hysteresis. The results from specimen to specimen were quite repeatable and could be said to define the path of passive motion of the human knee. During passive flexion to 90°, the femur rotates externally on a fixed tibia (or the tibia rotates internally on a fixed femur) through about 22°. There is also a small amount of abduction/adduction. The components of translation vary from point to point on the femur but all are uniquely coupled to flexion angle. All specimens exhibited just one degree of freedom.
Figure 3.5 Axial rotation (mean values ± one standard deviation) plotted against flexion angle for the 12 specimens of Fig. 3.4 (marked W) and for 10 further unloaded specimens when tested in a six-degree of freedom rig simulating deep squats with the hip moving vertically above the ankle (marked Z).
Figure 3.5 shows results obtained by Zavatsky from 10 specimens tested in a rig simulating deep squats during passive motion (O’Connor et al. 2003.; Biden & O’Connor 1990; Zavatsky 1997). The ‘Z’ curves are almost identical to the graphs of axial rotation of Figures 3.2, 3.3 and 3.4, demonstrating that the passive motion described by these figures is characteristic of the human knee and not of the apparatus in which the measurements are made. Such results can be achieved only if the apparatus used to hold the specimens allows them six independent degrees of freedom so that the only constraints applied to the motion of the specimens are the passive constraints of the ligaments and articular surfaces (Zavatsky 1997).
Blankevoort et al. (Blankevoort et al. 1988) did not find such a unique path but defined an envelope of external/internal rotation about 40° wide obtained when the knee was flexed first with an internally rotating torque then with an externally rotating torque of 3 Nm. We show below that ordered motion is observed also in the presence of torques of value ±1 and ±2 Nm (Fig. 3.22).
The highly ordered pattern of movement suggested by Figures 3.2, 3.3 and 3.4 was not repeated in three further specimens which showed disorderly movements and wide hysteresis loops (Fig. 3.6). On subsequent dissection, two of these specimens were found to have articular surfaces eroded by disease (Figs. 3.6(a) and 3.6(b)) and the third had partial division of the MCL (Fig. 3.6(c)). For each specimen, the path followed during flexion was very different to that followed in extension, and the angle of axial rotation and the other four degrees of freedom were no longer uniquely coupled to the flexion angle. These specimens, which no longer had any preferred path, did not resist being positioned anywhere between the upper and lower curves. These results demonstrate that the ordered movement exhibited by the human knee during passive movement requires fully intact articular surfaces and fully intact ligaments.