Predicting in vivo clinical performance of anterior cruciate ligament fixation methods from in vitro analysis: industrial tests of fatigue life and tolerance limits are more useful than other cyclic loading parameters.

Background: Anterior cruciate ligament graft fixations experience cyclic loads in vivo. Present cyclic loading studies testing fixation use "incremental cycling," "residual strength" protocols, or a combination. Industrial standards, however, rely on fatigue life testing and use tolerance limits to determine guaranteed minimum levels of cycles to failure.

Hypothesis: Industrial standards of fatigue life and lower tolerance limits provide a more conservative assessment of cycles to failure than do currently used cyclic loading models, and they facilitate interpretation of data toward clinical performance.

Study Design: Descriptive laboratory study.

Methods: Fatigue life curves and lower tolerance limits were calculated for porcine patellar tendon graft fixations that were tested to failure in single pulls or cycled to failure at 2 different load levels. A log curve was constructed so that the expected values at other load levels could be calculated. Comparison of a metal and a composite resorbable interference screw design was used as an example.

Results: Cycles to failure of both screw designs varied widely at each load level. The guaranteed minimum number of cycles calculated by tolerance limits was therefore much lower than the mean cycles to failure at any given load level. For example, at a load of 250 N, the predicted mean life for the composite screw and the metal screw was 2513 and 1490 cycles, respectively, whereas the 80/80 tolerance limits (the value that 80% could be expected to achieve with 80% confidence) were only 63 and 68 cycles, respectively. Small reductions in load level gave very large increases in minimally expected life.

Conclusions: Fatigue testing of anterior cruciate ligament reconstructions better shows the wide variation between specimens and the large effect of load levels on expected life. Wide scatter makes mean levels over-optimistic and difficult to interpret. Tolerance limits give estimates that are more conservative and facilitate data interpretation.

Clinical Relevance: A method for testing and analyzing fatigue properties was presented, results of which are more readily interpreted to clinical practice.