Knee prostheses of eight different designs were tested experimentally to determine the axial torque necessary to rotate the tibial component relative to the femoral component with the prosthesis positioned at or near full extension. The results represent transmitted torque as a function of axial rotation. For six designs, the transmitted torque was 11.3 newton-meters (100 inch-pounds) or greater at 6 degrees of rotation, with an axial load of 1556.8 newtons (350 pounds). These prostheses either had close conformation between femoral and tibial components, or were mechanically linked. In these prostheses, the geometry of the articulating surfaces was the primary determinant of torque. For the other two types, under similar loading conditions the transmitted torques were 2.5 newton-meters (twenty-two inch-pounds) at 6 degrees of rotation. These prostheses were designed to allow greater freedom for rotation, and hence for them the primary determinant of torque was friction between the articulating surfaces. The predictive value of this torque characteristic relative to the incidence of loosening will have to be determined by comparison with clinical experience.
Copyright 1978 by The Journal of Bone and Joint Surgery, Incorporated