The model patella used in our two-dimensional models (Gill & O’Connor, 1996a) is shown as a rectangle in Figures A3, A6, A9, and A10 with two straight-line articular surfaces parallel to its anterior surface. In addition to a circle defining the sulcus of the trochlea, Figures A7–A10 contain a more anterior curve outlining the flanges of the trochlea.
The most posterior surface of the model patella, representing the median ridge of the human patella, makes contact with the femoral trochlea at extension and over most of the flexion range. The more anterior surface represents the medial and lateral facets of the patella. Patellofemoral contact occurs near the distal pole of the patella in extension (Figs. A3(a) and A7(a)), and moves proximally (Figs. A3(b) and A7(b)) with increasing flexion. Eventually, contact moves onto the medial and lateral facets of the patella, and onto the femoral condyle (Fig. A3(c)) or the femoral component (Fig. A7(c)). The median ridge (the anterior articular surface) then passes into the intercondylar notch and lies proximal to the articular surface of the femur.
The predicted movement of the contact area on the model patella is very similar to that observed in the human knee (see Fig. 5.1) (Goodfellow et al., 1976). The patella rolls proximally as it slides distally on the femur during knee flexion, ensuring that the line of action of the patellofemoral contact force always passes through the intersection of the patellar and quadriceps tendons, a condition necessary for mechanical equilibrium of the patella (Gill & O’Connor, 1996a).
A number of the predictions of patellofemoral behaviour from this model agree well with independent measurements (O’Connor et al., 1990). The quadriceps tendon is shown as a line extending proximally from the proximal pole of the model patella (Figs. A3, A6, A7–A10). At high flexion angles (>85°), it wraps around the trochlea to form the tendofemoral joint, as described by Bishop and Denham (1977). The angle between the quadriceps and the patellar tendons, called the patella mechanism angle, diminishes with increasing knee flexion. The variation of this angle as calculated from the model agrees well with measurements by Buff et al. (1988). The angles between each of the tendons and the line of action of the patellofemoral force are not equal, and as a result the tension forces in the two tendons are not equal (Maquet, 1976). The variation of the ratio of the tendon forces with flexion angle calculated from the model agrees well with measurements (Bishop & Denham, 1977; Buff et al., 1988; Maquet, 1976; Ellis et al., 1984; Huberti et al., 1984).
The instant of transition from trochlear contact to condylar contact is shown in Figure A10, with the model knee drawn at 99° flexion. Contact is about to move from the trochlea onto the femoral component. The patella is held anteriorly on the trochlea until its medial and lateral facets overlie the femoral component. The anterior edge of the femoral component is bypassed and is never in contact with the patella, so that there is no danger of the type of impingement described by Hernigou and Deschamps (2002) (see Fig 5.2) as a common complication of fixed-bearing polycentric unicompartmental arthroplasties.
The models of the intact and replaced joints both demonstrate changes in PTA with flexion similar to those observed in living patients (see Fig. 3.27) with complete restoration of function. This may be a reason why revision of an OUKA for patellar problems is rare.
