Abstract
Introduction
Contemporary total knee systems accommodate for differential sizing between femoral and tibial components to allow surgeons to control soft tissue balancing and optimize rotation. One method some manufacturers use to allow differential sizing involves maintaining coronal articular congruency with a single radius of curvature throughout sizes while clipping the medial-lateral width, called a single coronal geometry system. Registry data show a 20% higher revision rate when the tibial component is smaller than the femur (downsizing) in the DePuy PFC system, a single coronal system, possibly from increased stresses from edge loading or varying articular congruency. We examined a different single coronal geometry knee system, Smith & Nephew Genesis II, to determine if edge loading is present in downsized tibial components by measuring area and location of deviation of the polyethylene articular surface damage.
Methods
45 Genesis II posterior-stabilized polyethylene inserts (12 matched and 33 downsized tibial components) were CT scanned. 3D reconstructions were registered to corresponding pristine component reconstructions, and 3D deviation maps of the retrieved articular surfaces relative to the pristine surfaces were created.
Each map was exported as a point cloud to a custom MATLAB code to calculate the area and weighted center of deviation of the articular surfaces. An iterative k-means clustering algorithm was used to isolate regions of deviation, and a shrink-wrap algorithm was applied to calculate their areas. The area of deviation was calculated as the sum of all regions of deviation and was normalized to the area of the articular surface. The location of deviation was described using the weighted center of deviation and the location of maximum deviation on the articular surfaces relative to the center of the post (Fig. 1).
Pearson product moment correlations were conducted to examine the correlation between length of implantation (LOI) and the medial and lateral areas of deviation for all specimens, matched components, and downsized components.
Results
The mean LOIs for downsized and matched tibial components were not different (36±28 months vs 46±26 months, p=0.24). Areas of deviation for the medial and lateral sides for both downsized and matched components increased with LOI (p<0.001). Medial and lateral sides of matched retrievals were not different in location of maximum deviation, maximum deviation, and weighted center of deviation (p>0.4). The matched and downsized retrievals did not have different centers of deviation in the medial-lateral direction, maximum deviations, or locations of maximum deviations (p>0.1).
Discussion
Our results suggest that downsizing the tibial component in the Genesis II system, a single coronal geometry system, did not affect the area or location of deviation on the articular surface. Overall, the weighted center of deviation remained close to the dwell point and did not change as a function of tibial downsizing. However, we saw deviation patterns biased peripherally for inserts with low LOI in both matched and downsized cohorts. With increasing LOI, the deviation expanded to cover the majority of the available articular surface. Our results suggest the need to further examine this and other systems determine the effects of differential sizing.
For figures/tables, please contact authors directly.
