Abstract
In replacing the human knee, we attempt to reproduce the stability of the normal knee so that the knee will feel as close to normal as possible to the patient. To answer the question, “Which features matter?” we must first examine the stability of the normal knee. Compliance and stiffness: Stability is measured as “force-displacement” behavior. That is, a force is applied to the knee and the relative motion is measured. Engineers refer to the curves generated by this type of experiment as “stiffness”. Because stiffness is not a term that orthopaedists like to hear when referring to a knee, the inverse term “compliance” often is used. Ligament stress-strain: The force-displacement test for ligaments is called a “stress-strain” curve and shows three regions of force-displacement response. Early in loading a small force causes considerable displacement. This is called the “toe region” of the curve. After a certain amount of displacement, the ligament enters the “elastic region” of the curve and becomes markedly more stiff. Finally, if enough force is applied, the ligament begins to fail at its “yield point”. Ligaments “live” in the toe region of the stress-strain curve. This can be seen clinically when, in response to varus-valgus and anteroposterior stress, the tibia moves relative to the femur until it is stopped by tension in the ligament. This is the ligament moving from the toe region into the elastic region. Compliance of the knee: In a number of studies done in the 1970s, the compliance of the knee was found to be least to both varus-valgus and anteroposterior loads in full extension. In flexion, compliance increases particularly to varus-valgus stress. This implies that the ligamentous structures about the knee are most tight in extension and become more lax in flexion. When external load is applied to the knee, either in the form of muscle contraction or bearing weight, the compliance of the knee decreases (i.e., it becomes more stiff and more stable). Loading will decrease the tension in the ligaments, yet the knee is less compliant. The only way this can happen is by the geometry of the surfaces imparting the stability. The conclusion from these studies is that the human knee, when moving in the usual plane of motion, is stabilised by the geometry of the surfaces, or the congruency of the femur and tibia. Ligaments are recruited to limit motion when forces outside the plane of motion (“out-of-plane” loads) are applied to the knee. These loads move the knee ligaments from the toe region into the elastic region of their stress-strain curve.
Two kinds of total knee prosthesis design: Most total knees are designed to have little or no congruence between the femur and tibia, likely because of the worry about “kinematic conflict” that dates to the four-bar-linkage model of knee motion first proposed by Zuppinger in 1907. In these types of total knees, the ligaments are tensioned (i.e., “balanced”) so that they do the job done in the normal knee by congruence. A few total knees are designed for congruence between the femur and tibia, either in just the medial compartment or in both compartments. The answer to the question, “What is needed for total knee stability?” For non-congruent knee prostheses, the ligaments must be balanced or tensioned into the elastic portion of the stress-strain curve so that the knee is stable. The ligaments must remain in the elastic region indefinitely or the knee will be unstable. For congruent knee prostheses, the ligaments can be left in the toe region and rely, similar to the normal knee, on the geometry of the surfaces to provide stability and allow the ligaments to be recruited for out-of-plane loads. The ligaments must not be left too loose, lest the knee be unstable to out-of-plane loads but must not be as tight as is done with ligament tensioning prostheses.
