25–40% of unicompartmental knee replacement (UKR) revisions are performed for unexplained pain possibly secondary to elevated proximal tibial bone strain. This study investigates the effect of tibial component metal backing and polyethylene thickness on cancellous bone strain in a finite element model (FEM) of a cemented fixed bearing medial UKR, validated using previously published acoustic emission data (AE). FEMs of composite tibiae implanted with an all-polyethylene tibial component (AP) and a metal backed one (MB) were created. Polyethylene of thickness 6–10mm in 2mm increments was loaded to a medial load of 2500N. The volume of cancellous bone exposed to <−3000 (pathological overloading) and <−7000 (failure limit) minimum principal (compressive) microstrain (µ∊) and >3000 and >7000 maximum principal (tensile) microstrain was measured. Linear regression analysis showed good correlation between measured AE hits and volume of cancellous bone elements with compressive strain <−3000µ∊: correlation coefficients (R= 0.947, R2 = 0.847), standard error of the estimate (12.6 AE hits) and percentage error (12.5%) (p<0.001). AP implants displayed greater cancellous bone strains than MB implants for all strain variables at all loads. Patterns of strain differed between implants: MB concentrations at the lateral edge; AP concentrations at the keel, peg and at the region of load application. AP implants had 2.2 (10mm) to 3.2 (6mm) times the volume of cancellous bone compressively strained <−7000µ∊ than the MB implants. Altering MB polyethylene insert thickness had no effect. We advocate using caution with all-polyethylene UKR implants especially in large or active patients where loads are higher

Unicompartmental knee replacements (UKRs) have inconsistent and variable survivorships reported in the literature. It has been suggested that many are revised for ongoing pain with no other mode of failure identified. Using a medial UKR with an all-polyethylene non-congruent tibial component from 2004–7, we noted a revision rate of 9/98 cases (9.2%) at a mean of 39 months. Subchondral sclerosis was noted under the tibial component in 3/9 revisions with well fixed implants, and the aim of this study was to investigate this as a mode of failure. 89 UKRs in 77 patients were investigated radiographically (at mean 50 months) and with SF-12 and Oxford Knee scores at mean follow up 55 months. Subjectively 23/89 cases (25%) had sclerosis under the tibial component. We describe a method of quantifying this sclerosis as a greyscale ratio (GSR), which was significantly correlated with presence/absence of sclerosis (p<0.001). Significant predictors of elevated GSR (increasing sclerosis) were female sex (p<0.001) and elevated BMI (P=0.010) on both univariate and multivariate analysis. In turn, elevated GSR was significantly associated with poorer improvement in OKS (p<0.05) at the time of final follow up. We hypothesise that this sclerosis results from repetitive microfracture and adaptive remodelling in the proximal tibia due to increased strain. Finite element analysis is required to investigate this further, but we suggest caution should be employed when considering all polyethylene UKR implants in older women and in those with BMI >35