The mobile bearing passively follows the track of the femoral condyle as it moves anteroposteriorly and mediolaterally relative to the plateau. The limits of these movements are set by the lengths of the ligaments, and, within these limits, the freedom of the bearing to translate in all directions should be unrestricted. If the bearing is unrestrained, loads are transmitted across the joint as a combination of compressive force (perpendicular to the articular surface) and tensile forces in the soft tissues. Shear stresses at the bone–implant interfaces are thereby minimised. If bearing movement is resisted (by impingement against bone, cement, or the lateral wall of the tibial component), shear and tensile stresses may develop at the femoral and/or tibial interfaces and could cause loosening of the fixed components, dislocation of the bearing and perhaps pain (see Fig. 2.20).
The largest translational movements of the femoral condyle relative to the tibia are in the anteroposterior direction. The flat plateau of the tibial component, having no anterior or posterior rim, offers no limits to movement of the bearing in these directions, nor do the soft tissues interfere. In flexion, when the bearing moves posteriorly, the posterior capsule is relaxed; in extension, when the bearing moves forward, so does the anterior capsule. In a normally functioning OUKA, the excursion of the bearing regularly brings its posterior margin beyond the posterior edge of the tibial plateau in flexion, and anteromedial overhang is common in full extension. Despite these movements of the bearing, the centre of pressure always lies within the middle third of the tibial implant, maintaining the bone/implant interface in compression.
In a knee with normally tensioned ligaments, the femoral condyle has very little freedom to translate mediolaterally, and the bearing needs equivalently little freedom to follow it. However, the bearing should not be jammed against the lateral wall of the tibial component.
Rotation and entrapment
As well as the translational movements described above, the femoral condyle rotates axially relative to the tibial plateau. Obligatory internal rotation of the femur, of about 20° (see Figs. 3.2 and 3.3(a)), occurs during passive extension, and a range of forced internal/external rotation is available in all flexed positions due to tissue deformation, increasing with increasing flexion (see Fig. 3.22). Axial rotation of the femur relative to the tibia is accomplished by means of spinning movements at the femoromeniscal interface and anteroposterior translational movements at the meniscotibial interface. Since the femoromeniscal interface of the prosthesis is a ball-in-socket and the meniscotibial interface is flat-on-flat, spinning movements can occur at both surfaces of the bearing. One of these levels of rotation is superfluous, and so spin can be limited at the lower interface, by the tibial wall, without limiting the joint’s freedom to rotate. The design takes advantage of this to maximise entrapment of the bearing.
The prototype bearing (1976 to 1978) was circular in plan and the amount of entrapment, i.e. the height of the socket wall above its deepest point, was the same all round (Fig. 6.13(a)). However, dislocation proved more likely to occur anteriorly or posteriorly than sideways and it was desirable to have more entrapment at the front and the back of the socket than at the sides (Goodfellow & O’Connor 1992). Furthermore, with the circular design, entrapment can only be increased by decreasing the radius of the femoral component or increasing the width of the bearing, and the latter would make the bearing too wide for the tibial plateau. The solution was to make the bearing quadrilateral in plan, with its longer axis in the anteroposterior direction (Fig. 6.13(b)). This allowed the anterior and posterior lips of the socket to be higher (increasing anteroposterior entrapment without affecting mediolateral entrapment).
Since it is difficult to stretch the ligaments more than 3 mm, the posterior wall of the socket cannot provide greater entrapment than that (or the bearing cannot be inserted). Therefore the centre of the socket was moved towards the back of the quadrilateral, diminishing the height of the posterior lip to about 3 mm and raising the anterior lip to about 5 mm (Fig. 6.13(c)). This is the general form that the medial bearing has had ever since the Phase 1 implant was introduced in 1978.
Figure 6.13 Bearing design and entrapment.
Figure 6.14 Limiting rotation of the bearing.
Obviously, this mechanism can only work if the bearing maintains its anteroposterior alignment. If it spins through 90°, the socket will present reduced resistance to anterior or posterior dislocation (Fig. 6.14(a)). Figure 6.14(b) shows how its quadrilateral shape limits spin of the bearing if the centre of the socket (about which spin occurs) is close enough to the lateral wall of the tibial implant. This is achieved by positioning the femoral component appropriately. Unlimited spin is still allowed at the meniscofemoral interface and the fundamental requirement, that the bearing be free to translate in all directions, is still satisfied.
In order to minimise further the risk of bearing rotation, ‘anatomic’ bearings (right- and left-sided) with an extended lateral edge were introduced in 2002/3 (Fig. 6.14(c)). Also, the anteromedial corner has been rounded off to minimise bearing overhang in extension and the potential for soft tissue irritation.