Phase 1
In 1974, two of the authors (JWG and JJOC) introduced congruous mobile bearings for knee prostheses (Goodfellow, O’Connor & Shrive, 1974). The first ‘Oxford Knee’ had a metal femoral component with a spherical articular surface, a metal tibial component which was flat, and a polyethylene mobile bearing, spherically concave above and flat below, interposed between them (Fig. 1.2). The device was fully congruent at both interfaces throughout the range of movement (to minimise polyethylene wear) and fully unconstrained (to allow unrestricted movements and minimise the risk of loosening). These features of the Oxford Knee have remained unchanged to the present day.
Figure 1.2 The Oxford Knee (Phase 1) (1978).
At first, the implant was used bicompartmentally, as a total joint replacement, with two sets of components inserted one medially and one laterally. The non-articular surface of the femoral component of the original design (Phase 1) had three inclined facets and was fitted to the femur by making three saw-cuts as shown in Figure 1.3. Many surgeons found it difficult to locate the femoral component accurately in relation to the ligaments and, therefore to match the extension gap to the flexion gap.
Figure 1.3 Method of preparation of the femur for the Oxford Knee (Phase 1).
It became apparent that good results were only achieved if the ACL was intact (Goodfellow & O’Connor, 1992). Another observation was made, that if the ACL was intact, then the arthritis tended to be confined to the anteromedial part of the tibia and the distal part of the medial femoral condyle. In these cases, all ligaments were functionally normal. This disease was called Anteromedial OA (AMOA) (White et al., 1991). On the basis of these two observations, in 1982 the device began to be used unicompartmentally and the primary indication was AMOA.
Phase 2
In 1987, the Phase 2 implant was introduced specifically for unicompartmental arthroplasty (White et al., 1991). The non-articular surfaces of the femoral component had a flat posterior facet and a spherically concave inferior facet (Fig. 1.4). The posterior femoral condyle was prepared by a saw-cut and its inferior facet was milled by a spherically concave bone-mill rotating round a spigot in a drill- hole in the condyle (Fig. 1.5). By shortening the spigot, measured thicknesses of bone could be milled incrementally from the inferior surface of the condyle, allowing the gaps in flexion and extension to be balanced intraoperatively and simultaneously shaping the bone to fit the implant (Goodfellow & O’Connor, 1989). This accurate system for restoring ligament tension to normal not only decreased the bearing dislocation rate to very low levels but also restored knee kinematics, and thus function, to normal. The 10 year results of the Phase 1 and 2 Oxford Knee were published by the designers in 1998 (Murray et al., 1998) and by an independent surgeon, Dr Svard in 2001 (Svard & Price, 2001). These two studies demonstrated, for the first time, that the long term survival of UKA can be as good as that achieved by TKA. Similar results were also shown by Barrington and Emerson in 2010 (Barrington & Emerson, 2010).
Figure 1.4 The Oxford Unicompartmental Knee Arthroplasty (OUKA (Phase 2)) (1987).
Figure 1.5 Method of preparation of the femur for the OUKA (Phase 2) with a concave rotary mill turning around an adjustable spigot.
Phase 3
The Phase 1 and Phase 2 prostheses were implanted through an open approach with dislocation of the patella, as in TKA. In 1998, the Phase 3 prosthesis was introduced specifically for medial unicompartmental use with a minimally invasive approach (Fig. 1.6). The single size of femoral component (used in all the Phase 1 and 2 implants) was replaced by five parametric sizes, the universal tibial plateau was replaced by right- and left-handed tibial components and the bearings were modified to diminish the likelihood of impingement and rotation. The instruments were miniaturised to facilitate their use through a small parapatellar arthrotomy. The functional results and speed of recovery of Phase 3 were found to be better than those of Phase 2 (Price et al., 2001).
Figure 1.6 The OUKA (Phase 3)(1998) implanted through a small incision.
In 2004, cementless components based on the Phase 3 were first used (Fig 1.7). Apart from having a porous titanium and hydroxyapatite coating, the main difference between the cementless and the original Phase 3 was that the cementless femoral component had two pegs. Because, generally, cementless TKA have not performed as well as cemented (Forsythe et al.. 2000), the cementless OUKA was carefully assessed. Randomised controlled trials with clinical and Radiostereometric (RSA) outcome measures were used to assess fixation, and a large multicentre cohort study was used to assess complications and contraindications (Liddle et al., 2013). Because the results of cementless have proved to be at least as good as those of cemented, many experienced surgeons now just use the cementless OUKA.
It was clear from the cementless experience that a two peg femoral component was preferable to a single peg. Therefore, a two peg cemented component, that was similar to the cementless, was introduced (Fig 1.8). A clinical study has shown that the two peg cemented component performs well (White et al., 2012).