Les montages les plus fréquemment utilisés pour l'analyse cinématique du genou représentent en fait une succession d'états statiques. Les auteurs ont développé un montage original, mobilisant de façon continue et coordonnée fémur et tibia par des actionneurs asservis informatiquement, permettant notamment d'analyser la cinématique fémoropatellaire tridimensionnelle.

Cinq paires de pièces cadavériques fraîchement congelées, conservant les structures capsuloligamentaires du genou et le tendon du quadriceps ont été étudiées. Le mouvement simulé représentait le passage de la position debout (angle sagittal fémorotibial de 180°, angle sagittal tibia-sol de 90°) à la position assise (angle sagittal fémorotibial de 90°, angle sagittal tibia-sol de 45°).

Neuf marqueurs réfléchissants répartis en trois tripodes étaient ancrés dans la corticale du fémur, du tibia et de la patella, et leurs déplacements mesurés par trois caméras infrarouges. Les mouvements respectifs des pièces osseuses dans les trois plans de rotation et les trois plans de translation ont été interprétés sous forme de courbes en fonction de l'angle de flexion du genou.

La patella du genou sain effectuait un mouvement de flexion continue par rapport au fémur, de rotation interne en début de flexion puis de rotation externe par rapport au fémur, de translation antérieure et de translation médiale par rapport au tibia. L'implantation d'une prothèse unicompartimentale médiale n'a provoqué aucune modification de la cinématique fémoropatellaire. L'implantation d'une prothèse totale (prothèse Search avec implantation rotulienne en dôme) a modifié sensiblement la cinématique fémoropatellaire, avec notamment une rotation externe parasite induite par les modifications de la cinématique fémorotibiale en rotation. Aucune influence de l'état du pivot central n'a été retrouvée. La cinématique fémoropatellaire était donc essentiellement imposée par la géométrie prothétique.

In vitro experiments are particularly useful for studying kinematic changes from the normal knee to experimental conditions simulating different disease states. We developed an experimental protocol allowing a kinematic analysis of the femorotibial and femoropatellar joints in the healthy knee and after implantation of a knee prosthesis, according to the central pivot during simulated active loaded movement from the standing to sitting position.

An experimental device was designed to apply force to the femur of a cadaveric specimen including the femur, the patella and the tibia. The tibia was angled in the sagittal plane and the femur was free to move in space in response to the geometric movement of the knee joint, the capsuloligamentary structures, the quadriceps tendon and gravity. Variation in the length of the quadriceps tendon controlled the flexion-extension movement. The experimental setup included computer-controlled activation allowing continuous coordinated movement of the femur relative to the tibia and of the tibia relative to the ground. Standard activations simulated movement from the standing to the sitting position.

Five pairs of fresh-frozen cadaver specimens including the entire femur, patella, tibia and fibula, the capsuloligamentary and intra-articular structures of the knee, the superior and inferior tibiofibular ligaments and the quadriceps tendon were studied. The quadriceps tendon was connected to the computer-guided activation device. Reflectors were fixed onto the anterior aspect of the femur, the superior tibial epiphysis and the center of the patella. Anatomic landmarks on the femur, the tibia, and the patella were identified to determine the plane of movement of each bone in the three rotation axes and the three translation directions. Three infrared cameras recorded movements of the reflectors fixed on the bony segments and, by mathematical transformation, the movement of the corresponding bony segment, displayed in time-course curves.

The patella moved in continuous fashion over the femur, directly following the angle of knee flexion with a ratio of about 60%, which was constant for all knees studied and for all configurations. The patella of healthy knees and knees implanted with a unicompartmental prosthesis exhibited medial rotation during the first 30° of flexion, with a movement of about of 10°, then a lateral rotation of about 10° to 20° when the flexion reached 90°; implantation of a total knee prosthesis led to a medial rotation which was continuous from 5° to 15°. There was a trend towards continuous abduction of about 10°. The patella exhibited a continuous anterior translation of 10 to 20 mm from the tibia with increasing knee flexion, in both normal and prosthetic knees (unicompartmental prosthesis); knees implanted with a total knee prosthesis exhibited 5 to 10 mm anterior translation from 0° to 50° flexion, then an equivalent posterior translation for 50° to 90° flexion. The patella made a continuous 5 to 10 mm medial translation movement over the tibia in both normal and prosthetic (unicompartmental) knees; knees implanted with a total knee prosthesis exhibited 0 to 5 mm lateral translation starting after 50° flexion. The patella also exhibited a continuous distal translation over the tibia of about 20 to 30 mm, for all configurations.

The experimental set up enables a comparison of the kinetics of a normal knee with the kinetics observed after implantation of a prosthesis on the same knee. Implantation of a unicompartmental medial prosthesis, leaving the posterior cruciate ligament intact and irrespective of the status of the anterior cruciate ligament, did not, in these experimental conditions, exhibit any significant difference in the femorotibial or femoropatellar kinetics compared with the same normal knee.

Implantation of a total knee prosthesis had a significant effect on the femoropatellar kinematics, compared with the same knee before implantation. The main anomalies were related to the medial-lateral rotation of the patella which exhibited an abnormal lateral rotation, possibly favorable for subluxation; these changes were directly related to femorotibial rotation after implantation of the total prosthesis and appeared to be related to the symmetry of the femoral condyles of the prosthesis model studied, perturbing the normal automatic rotation of the knee. There is thus a strong relationship between femorotibial and femoropatellar kinetics in the total knee prosthesis.