The 2014 NJR report gives a figure of 0.30 revisions for periprosthetic fracture per 1000 years (95% CI 0.24 – 0.37) in UKA (Fig. 11.3(a)) (The NJR Board, 2014). Registers however underestimate the incidence of fracture as many are internally fixed rather than revised. In our series of 1000 Phase 3 cemented OUKAs, we did not encounter a single fracture (Pandit et al, 2011). Similarly with the current design of cementless OUKA, we have not encountered any tibial plateau fracture. Fractures occur with all types of UKA, for example four plateau fractures occurred in 62 UKAs reported by Berger et al. (1999) and there are other occasional reports in the literature (The NJR Board, 2014; Swanson et al, 1985; Yang et al, 2003; Pandit et al, 2007; Rudol et al, 2007). We have reported on a series of eight cases of tibial plateau fractures after Oxford UKA (which occurred over a period of seven years) collected from various institutions in the UK (Pandit et al, 2007). The study confirmed that these fractures are rare and tend to occur in inexperienced hands with both cemented as well as cementless implants.
Despite the very few fractures in the published reports of OUKA cohorts, several have been reported to us by other users and we believe that the complication is more common than the literature suggests, particularly among surgeons beginning to do unicompartmental arthroplasty, in populations with constitutionally small tibiae (e.g. Asian populations) and following cementless fixation.
Cause
It seems likely that plateau fractures either occur intraoperatively or are a result of severe weakening of the bone caused intraoperatively. They are either diagnosed intraoperatively or present later, commonly at 2–12 weeks postoperatively (Fig. 11.3(a)). If the fracture is initially undisplaced, it may not be visible on the immediate postoperative radiographs, only appearing later when weight-bearing has caused displacement, and postoperative pain and deformity have drawn attention to the problem.
Weakening of the condyle
Weakening of the condyle by removal of its articular surface and subchondral bone plate is probably the main reason for fracture. Since this is unavoidable, great care should be taken to avoid any additional weakening of the bone. We believe that the most potent cause of fracture is damage to the posterior tibial cortex and the cancellous bone by vertical saw cuts that go deeper than they need or if the keel slot breaks through the posterior cortex (Fig. 11.3(c)). Even the two small holes made in the anterior cortex by the nails that fix the tibial saw guide have been shown to decrease the strength of the tibial condyle (The NJR Board, 2014; Swanson et al, 1985). Although we have used this method of fixation in all phases of the OUKA without complication, we now use only one nail.
Figure 11.3 (a) Fracture of the medial tibial condyle. (b) The bearing does not dislocate because the width of the ‘gap’ is maintained by the intact MCL. (c) The probable cause of fracture is the medially placed vertical cut and deep keel saw cuts made during preparation of the groove for the tibial keel. In this case, the cuts (arrowed) are about twice the depth of the tibial keel.
The more bone that is removed from the condyle, the weaker is the remainder; therefore the surgical objective should be to remove as little as possible. It is an advantage of meniscal-bearing arthroplasty that polyethylene only 3 or 4 mm thick can be used safely. This advantage should be exploited by removing as little tibial plateau as possible.
The smaller the tibia, the less bone can be safely removed from it. This may explain why more fractures have been reported to us from Asian countries where many adults are of small stature. It is particularly important in small patients that the horizontal tibial saw cut should aim to accommodate a thin bearing which can be as little as 3 mm thick (see Chapter 6). Cadaver studies have shown that the load to fracture is significantly lower following cementless fixation than cemented so surgeons must be more careful with cementless fixation (Seeger et al, 2012). This is a particular problem when surgeons with a large experience of the more forgiving cemented fixation swap to cementless.
Finite element analysis (FEA) work shows that the strain in the proximal tibia and thus the risk of fractures is significantly elevated by (a) a deep tibial resection; (b) too medial or too deep a vertical cut (Fig. 11.4) (Pegg et al, 2015). This in turn diminishes the area available for load transfer and, if the vertical cut violates the posterior cortex and/or the keel cut is made too deep, then a fracture may occur. The surgeon must not lift, and therefore tilt, the saw when performing the vertical cut. The vertical cut should be made just medial to the apex of the medial tibial spine. A simple way to ensure the vertical cut does not go too deep is to do the horizontal cut first and then insert a shim into the cut to act as an end stop for the vertical cut. (Extending the horizontal cut laterally tends to decrease the risk of fracture (Fig. 11.4).) The surgeon should also use the keel-cut saw blade and hold onto the pin fixing the template when performing the keel cut, to prevent the cut going too far posterior.
Figure 11.4 Finite element models showing the high risk of fracture (gray) following excessively deep vertical cuts (Pegg et al, 2015).
Application of excessive force may also cause fracture. The heavy hammer commonly used in TKA is not appropriate for UKA. A small hammer, in the hands of a surgeon alert to the risk, will seldom cause a tibial plateau fracture. This is a particular risk with cementless fixation as often the tibial plateau, after insertion, is left about 0.5 mm proud. This should be accepted as it will subside with time and integrate. Impacting it hard may cause a fracture.
Pathological anatomy
Figure 11.3(b) shows how the strong attachments of the MCL to the cortex of the medial tibial plateau maintain the components of the arthroplasty in their proper relationship. In particular, the gap between the femoral and tibial implants remains unaltered, and dislocation of the bearing does not occur. Displacement of the fragment distally allows the tibiofemoral axis to drift into varus.
