It is only in the laboratory, using a rig with six degrees of freedom and the weight of the bones counterbalanced, that it is possible to be sure of moving the knee without applying load to stretch its ligaments or indent its articular surfaces (Biden & O’Connor 1990). However, passive movements performed by the clinician while supporting the limb, particularly in anaesthetised subjects without muscle tone, can be similar to the unloaded state of the laboratory preparation.
In 1978, we reported that, in cadaver specimens with bicompartmental Oxford implants, forward movement of the meniscal bearings (from the rolled back position) was essential to the movement of extension and that, if bearing movement was blocked, extension was also blocked (Goodfellow & O’Connor, 1978).
Bearing and condyle movements in four unloaded cadaver knees flexed and extended in our deep squats rig (Biden & O’Connor 1990) after bicompartmental replacement were measured using a depth micrometer (Goodfellow & O’Connor, 1986). The medial bearings and condyles moved on average 12.5 mm backwards on the tibial component during flexion to 90° whereas the lateral bearings moved an average of 15.1 mm. Bearing movements of this order of magnitude are routinely observed on the operating table before closure. Bearing movements have also been measured during passive extension/flexion using fluoroscopy in nine patients with medial OUKA (a) after wound closure while they were still anaesthetised and (b) six months later in the follow-up clinic (Pegg et al., 2015). They reported a total backwards movement in the anaesthetised patients of 13 mm (SD 3 mm) during flexion to 120°, with differences between the anaesthetised and awake states only near extension (Fig. 3.16). Thus, roll-back in the replaced joint is accommodated by sliding at both the femoro-bearing and tibio-bearing interfaces. Although the total sliding distance (on both upper and lower interfaces) is greater than it would be in a fixed bearing implant, our evidence presented above in Chapter 2 demonstrates that full congruity between the components ensures minimal wear.
Figure 3.16. Bearing and condyle position on the tibial plateau in nine patients while anaesthetised and conscious during passive flexion/extension (Pegg et al., 2015).
Bradley et al. (1987) compared lateral radiographs of the knee in extension and in 90° flexion up to 5 years after OUKA (medial or lateral). The films were taken with the subjects lying on their side on the X-ray table with the muscles as relaxed as possible. The mean position of the bearings and condyles in flexion was 4.4 mm (range 0–13.5) posterior to their position in extension medially and 6.0 mm (range 1.6–13.0) laterally. The differences between these observations and those just described were attributed to the presence of passive muscle forces during the radiographic investigation.
Discussion
In the anaesthetised patient during passive extension/flexion, the only forces available to thrust the medial bearing forwards are the compressive forces at the articular surfaces and the tensile forces in the ligaments, both engendered by the force that the examiner applies in attempting to extend the leg or to support the flexing leg.
The movements of the bearings in the prosthetic joint, immediately after implantation and without muscle tone, were similar to the movements of the contact areas in the unloaded natural cadaveric knee (Table 3.1 above). The radiographic study at a longer follow-up time showed that bearings continue to move in the same sense for at least 5 years after implantation. However, the mean movement in that study was about half that seen intraoperatively and the spread of the data was greater; the differences were attributed to the presence of passive muscle tensions in the conscious patient. Bearing and condyle movements accommodate roll-back of the femur on the tibia during passive flexion.
The mathematical model of passive flexion of the replaced knee described in the Appendix and shown in Figure 3.17 explains the observations made on the unloaded prosthetic joint. At each flexion angle, the anteroposterior position of the femur on the tibia was determined from the condition of zero force in each ligament, a condition which required a roll-back of the model femur on the tibia of 7.5 mm over 90° of flexion. At each flexion angle, if the femur is placed anterior to this neutral position, the ACL (and MCL) are stretched and the PCL and LCL slackened (Fig. A8 in the appendix below), requiring the application of external force. If the femur is placed posterior to the neutral position, the PCL and LCL are stretched and the ACL and MCL are slackened. Thus, roll-back during passive flexion of the replaced joint is required to ensure that passive structures of the joint, articular surfaces as well as ligaments, are not loaded.
Figure 3.17. Model of the replaced knee in extension, 60° and 120° flexion, with a circular femoral component, a flat tibial component and an interposed fully conforming meniscal bearing. The model cruciate ligaments of Fig. 3.15 are also shown (see website, Animation 7).
Note that the shape changes and patterns of slackening and tightening of the ligaments in the model replaced joint are very similar to those of the model intact joint (Fig. 3.15), and those of the cadaver joint sketched by Friederich et al. (1992) (Fig. 3.12). For this reason, the Oxford implant can be said to be a ligament compatible prosthesis. Although the articular surfaces of the prosthesis are not exactly anatomical, they are sufficiently compatible with the retained ligaments to allow close restoration of natural passive motion. Calculations with the model of the replaced joint suggest that the most anterior fibres of the ACL and of the posterior bundle of the PCL remain isometric to within 0.5% during passive flexion/extension. Restoration of natural passive and dynamic laxity, and of normal kinematics and mechanics, might therefore be expected. Of course, restoration of natural laxity requires that the ligaments are restored to their natural strain patterns. This is achieved by matching the thickness of the gaps between the metal components at 110° and at 20° flexion, and filling the matched gap with a meniscal bearing of appropriate thickness, as explained in Chapters 6 and 7 below.
Note also that the model patella rolls on the anterior femur, making trochlear contact in Figures 3.17(a) and (b) and contact with the femoral component in Figure 3.17(c). This changing pattern of patellar contact is very similar to the pattern of contact in the intact knee (Figure 5.1 below). In our model, the transition from trochlear contact to component contact happens at 99° flexion, as shown in Figure 3.18. The figure shows the posterior surface of the patella (representing the central ridge) in contact with the trochlea and, simultaneously, the more anterior contact surface (representing the medial facet) in contact with the metal femoral component. The patella does not have to make contact with the hiatus between the trochlear flange and the implant and this may well be a reason for the very few revisions of the arthroplasty associated with patellar problems (Chapters 5 and 10).