The current design of cementless components was first used in 2004. It was important to assess accurately the results of these components before they were widely used because cementless total knee replacements have not performed as well as cemented total knees. The planned assessment included a randomised controlled trial based on RSA to compare the migration with cemented and cementless implants. In addition, another RCT was undertaken to compare the fixation based on screened radiographs of the cemented and cementless implants together with clinical scores. A large multicentre cohort study was undertaken because these two randomised studies were relatively small and therefore could not identify complications or contraindications. Finally, the outcomes will be assessed in the Joint Registries.
We used a model-based RSA system to compare the migration of cemented and cementless Oxford UKA (Kendrick et al, 2015). 43 knees were randomised. In RSA studies, the best predictor of long term loosening is increased migration in the second year. During the first year, both the cemented and cementless femoral components migrated 0.2 mm anteriorly and proximally. There was no significant difference between cemented and cementless. In the second year, there was no significant migration of either the cementless or cemented femoral components. During the first year, the cementless tibial components subsided significantly more than the cemented (0.3 mm v 0.1 mm p<0.01). However, during the second year, there was no significant difference in subsidence and both subsided 0.05 mm. The conclusion of the RSA study was that the cementless fixation was as good as the cemented.
In the clinical randomised study, 63 knees were randomised. Pre-operatively, there were no significant differences between the groups. The cementless procedure was 9 minutes quicker on average than the cemented. At 5 years, there was no significant difference between the OKS (cemented 39 (SD 10.4) and cementless 39.4 (SD 9.9)) and the KSS (Objective) (cemented 80.1 (SD 19.3) and cementless 78.8 (SD 14.0)). However, the KSS (Functional) was significantly better (p<0.01) for the cementless, (cemented 78.8 (SD 18.4) and cementless 92.0 (SD 12.7)). At 5 years, there was no significant difference in Tegner score although, at 2 years, the cementless was significantly better (p=0.04). There were no revisions.
The patients were assessed with fluoroscopically aligned radiographs. On the femoral side, there was no evidence of radiolucency or migration and no loosening. On the tibial side, there was also no evidence of loosening with no pathological radiolucencies or subsidence. However, there were physiological radiolucencies which were all less than or equal to 1 mm thick with a sclerotic margin (Kendrick et al, 2015). With cemented components, 25% had complete radiolucent lines, 36% had partial radiolucent lines and 35% had no radiolucent lines. The cementless components had significantly fewer radiolucent lines (p<0.001), with no complete radiolucencies, 7% partial and 93% no radiolucencies (Fig. 10.14).
Figure 10.14 Radiographs demonstrating radiolucent lines under (a) a cemented tibial tray and their absence (b) with a cementless implant.
Figure 10.15 Progression of radiolucencies in (a) cemented and (b) cementless OUKA in the RCT.
In the cemented group, there were no radiolucencies on the postoperative radiographs but they appeared during the first year. In contrast, in the cementless group, there were some radiolucencies immediately postoperatively, suggesting the components were not fully seated, but these disappeared within the first year. This study therefore suggests that the fixation of the cementless was better than cemented. It also suggests that the appearance of the interfaces did not change after a year so a one year follow up should give a good indication of the quality of fixation. The functional outcome of the cementless was as good, if not better than, the cemented.
A multicentre prospective study of cementless OUKA was undertaken in Oxford, Belfast (Northern Ireland) and Christchurch (New Zealand) (Naudie et al, 2004). 1000 knees (881 patients) were reviewed with a minimum follow-up of one year and a maximum of eight years. All patients had the standard indications used for cemented OUKA and the standard operative technique and rehabilitation regime. There were 19 (1.9%) reoperations. These consisted of dislocations (0.6%), OA progression (0.5%), peri-operative tibial fractures (0.3%), infections (0.3%), a late fracture from trauma (0.1%) and lateral AVN (0.1%). The seven year survival, when all reoperations were considered to be failures, was 97% which was similar to the cemented. The average pre-operative OKS was 21 (SD 7.9) and at one year was 40 (SD 7.8). 6.3% had partial radiolucencies and the remainder had no radiolucencies. There were no complete radiolucencies. No subgroup did badly suggesting that there were no additional contraindications for cementless compared with cemented. The conclusion therefore is that complication rate, survival and outcome for the cementless was similar to that of cemented. As there were no complete radiolucencies at a year, this is suggestive of improved fixation compared to the cemented.
Data from both the NJR and the New Zealand Joint Registry would suggest that the results of the cementless implant are better than the cemented. In the NJR, the five year survival of all Oxford arthroplasties was 93% (CI 92.7 – 93.5%, n = 33,272) and the five year survival of the cementless was 97% (CI 95.2 – 98%, n = 2,011). In New Zealand, the five year survival of the cemented implant was 93% (n = 3,267) and the cementless was 98% (n = 795). In both registers, the five year revision rate of the cementless arthroplasty was less than half that of the cemented and the difference was highly statistically significant. Although this does suggest the results of the cementless are better than the cemented, the better results may at least in part be explained by the fact that more experienced surgeons tend to use the cementless and also tend to have better results. A matched analysis of cemented and cementless components based on the NJR found that the OKS was significantly better with cementless (37.9 with cemented, 40.2 with cementless) (Liddle, 2014).