INTRODUCTION

It is widely accepted that the ligaments of the knee do not behave as homogeneous structures. This is especially important to the study of ligament injury mechanics since failure will begin in the most highly strained fibers. If the location of the highest strain varies as the knee moves through it's range of motion, then the failure mechanics could be expected to vary also. The current research project was undertaken to determine the variation in strain between two bundles of the ACL under the application of axial moments and quadriceps force.

REVIEW AND THEORY

Axial rotations of the tibia on the femur have been implicated in ACL ruptures during Alpine skiing. Internal axial moment at high flexion angles (e.g., Ettlinger, 1989) and external axial moment near full extension (e.g., Fischer et al. 1994) may cause ACL failure. Axial moments may also occur in conjunction with quadriceps force in some falls.

Numerous studies have examined the effects of axial moment (e.g., Pope et al., 1990), and muscle loads (e.g., Renstrom et al., 1986) on ACL strain. Unfortunately the loads used in the majority of these studies were insufficient for generating an injury model. Berns et al. (1993) attempted to develop such a model and found that the load-strain relationships could not be extrapolated to loads beyond the levels applied in the experiments.

Another potential complication to an injury prediction model might be the need to include terms for the strain in multiple sites in the ACL. Several studies have shown that the anteromedial (AMB) and posterolateral (PLB) bundles reciprocate in function over the range of flexion/extension (F/E). Unfortunately none of these studies examined the effects of applied loads on ligament strain. It could be expected that the site of maximum strain, and therefore the onset of injury, varies with applied load.

The objective of this study was to experimentally test the hypothesis that AMB and PLB strains are significantly different from each other. If the hypothesis was true, then an injury model would have to consider strain in at least two sites in the ACL. If it was false, then a single model for the entire ACL might be adequate.

METHODS

Liquid mercury strain gages (LMSGs) were installed on the surface of both the AMB and PLB of 6 unembalmed cadaver knee specimens (mean age 46.8 ( 16.5 years). The specimens were installed and aligned in a six-degree-of-freedom load application system (LAS) (Bach and Hull, 1995). This apparatus was pneumatically actuated and under full closed loop control. Each degree of freedom was individually instrumented for load and displacement measurement.

The specimens were subjected to a detailed loading protocol to examine the effects of axial and quadriceps loads, and combinations thereof, on the strain in the ACL. For combined loadings the quadriceps force was applied to the desired level followed by the axial moment. Loads were applied at knee flexion angles of 15, 30, 60, 90, and 120 degrees.

The data from these experiments were subject to a repeated measures analysis of variance (RANOVA) procedure. For this analysis three within-subjects effects were modeled and all possible interactions were considered. The 25 load levels analyzed involved permutations of 5 levels each (-10, -5, 0, 5, 10 Nm) of I/E moment applied in conjunction with 5 levels (0, 250, 500, 750, and 1000 N) of quadriceps force.

RESULTS

From the RANOVA results for I/E loading, it could be seen that only the flexion angle and load effects and their interaction were statistically significant (p=0.0005, p=0.0001, p=0.0008). The bundle effect was not significant (p=0.6080) nor was the flexion angle * load * bundle interaction (p=0.6732), although both the flexion angle * bundle (p=0.0853) and load * bundle (p=0.0986) interaction terms would have been significant at the (=0.10 level.

Though no significant differences between the two bundles were detected by the RANOVA, qualitative differences could be observed in the graphs of strain vs. internal and external axial moments (Figure 1 & Figure 2, respectively). Internal axial moment produced similar effects on strain relative to passive for both bundles (Figure 1). The effects of adding a quadriceps force produced similar effects on the strain due to an internal axial moment in the both bundles (Figure 1).

Figure 1 : Effects of internal axial moment applied with and without quadriceps force on ACL strain.

Figure 2 : Effects of external axial moment applied with and without quadriceps force on ACL strain.

External axial moment produced different results in the two bundles (Figure 2). For the AMB the application of an isolated 10 Nm external moment decreased the strain relative to passive for all flexion angles whereas for the PLB there was no effect on strain. For flexion angles of 15 to 60 degrees the external moment+quadriceps force combination increased strain in both bundles relative to passive strains (Figure 2).

DISCUSSION

Generating a model relating ligament strain to external and muscular loads as well as flexion angle is important to reducing the incidence of knee injuries in Alpine skiing. Such a model could be used in a programmable ski binding to monitor the state of strain based upon the state of these variables, and release the boot from the ski if the predicted strains approach injury levels. The ability to represent the ACL as a single structure rather than as multiple bundles would simplify both the experiments needed to generate the information for this model and the system which controls binding release.

The PLB strains due to the application of external moment + quadriceps force were visibly higher than those of the AMB. Between 15 and 30 degrees of flexion the PLB strain was up to 4% greater than the strain in the AMB. Although this difference was not statistically significantly, it can mean the difference between being injured or not being injured. This finding would indicate that although it might not be necessary to consider both bundles in an injury model, the model should be based upon the PLB.

REFERENCES

Bach, J.M.; Hull, M.L.; J. of Biomech. Eng., 117(4), 373-382, 1995.

Berns, G.S. et al. Skiing Trauma and Safety: Ninth Int. Symp., 89-110, 1993.

Ettlinger, C; Skiing, 85-121, 1989.

Fischer, J.F. et al., Acta Orth. Bel., 60(2), 194-203, 1994.

Pope, M.H. et al., Trans. of the First World Cong. of Biomech., 320, 1990.

Renstrom, P. et al., Am. J. .Sports Med., 14(1), 83-87, 1986.

ACKNOWLEDGEMENTS

The authors wish to thank the Tyrolia Corporation of Vienna, Austria for its continued support and Dr. Pat Patterson for his assistance with specimen preparation.