Load tolerance, security, and failure modes of fixation devices for synthetic knee ligaments.

One weak link in synthetic ligament reconstruction of the knee is the fixation device-ligament-bone interface. In this study, a laboratory model using bovine calf bone was designed to evaluate six different fixation devices combined with either an unlooped ligament in different configurations or a looped ligament. Results are expressed in terms of maximum load to failure, elongation at the site of fixation, stiffness of fixation prior to failure, and mode of failure. When failure of the test model occurred, it was always at the ligament--fixation device interface, the fixation device--bone interface, or a combination of the two. Elongation was recorded in all modes of fixation. The most secure fixation was obtained with a cortical screw and a looped ligament, the strength and predictability being mainly dependent on bone quality. The least secure and least predictable fixation was obtained with a prototype called bucket wedge. A barbed ligament staple and an unlooped ligament also had low predictability and considerably less strength than the screw fixation but showed less sensitivity to alterations in bone quality.