 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1997 The Journal of Bone and Joint Surgery, Inc. Biomechanical Evaluation of the Medial Collateral Ligament of the Elbow*   , NEW YORK, N.Y.Investigation performed at The Hospital for Special Surgery, Affiliated with The New York Hospital—Cornell University Medical College, New York City
Anatomical dissection and biomechanical testing were used to study twenty-eight cadaveric elbows in order to determine the role of the medial collateral ligament under valgus loading. The medial collateral ligament was composed of anterior, posterior, and occasionally transverse bundles. The anterior bundle was, in turn, composed of anterior and posterior bands that tightened in reciprocal fashion as the elbow was flexed and extended. Sequential cutting of the ligament was performed while rotation caused by valgus torque was measured. The anterior band of the anterior bundle was the primary restraint to valgus rotation at 30, 60, and 90 degrees of flexion and was a co-primary restraint at 120 degrees of flexion. The posterior band of the anterior bundle was a co-primary restraint at 120 degrees of flexion and a secondary restraint at 30 and 90 degrees of flexion. The posterior bundle was a secondary restraint at 30 degrees only. The reciprocal anterior and posterior bands have distinct biomechanical roles and theoretically may be injured separately. The anterior band was more vulnerable to valgus overload when the elbow was extended, whereas the posterior band was more vulnerable when the elbow was flexed. The posterior bundle was not vulnerable to valgus overload unless the anterior bundle was completely disrupted. The intact elbows rotated a mean of 3.6 degrees between the neutral position and the two-newton-meter valgus torque position. Cutting of the entire anterior bundle caused an additional 3.2 degrees of rotation at 90 degrees of flexion, where the effect was greatest. CLINICAL RELEVANCE: Physical findings in a patient who has an injury of the anterior bundle may be subtle, and an examination should be performed with the elbow in 90 degrees of flexion for greatest sensitivity. As the anterior bundle is the major restraint to valgus rotation, reconstructive procedures should focus on anatomical reproduction of that structure. Parallel limbs of tendon graft placed from the inferior aspect of the medial epicondyle to the area of the sublimis tubercle will simulate the reciprocal bands of the anterior bundle. Temporary immobilization with the elbow in flexion may relax the critically important anterior band of the reconstruction during healing.
The medial collateral ligament of the elbow is composed of an anterior bundle; a posterior bundle; and, occasionally, transverse fibers. The anterior bundle has been shown to be the primary restraint to valgus torque at the elbow9,16,18,25,28. Injury of the anterior bundle can cause subtle instability of the elbow and disabling pain in throwing athletes4,8,14,30,31. Complete tears of the anterior bundle of the medial collateral ligament have been described previously6,10,12,13,19. However, clinical examination of the elbow is difficult, and ancillary tests are of limited use in most cases. To improve the effectiveness of physical examination, we sought to determine the angle of flexion of the elbow at which excessive valgus angulation was most pronounced and was likely to be detected on manual stress-testing. We have frequently observed partial tears of the anterior bundle on magnetic resonance images and at the time of operative exploration. We believe that there are predictable patterns of injury of the anterior bundle and that these patterns are determined by the angle of flexion of the elbow at the time of injury. Previous authors have reported that the anterior bundle is composed of anterior and posterior bands that tighten in reciprocal fashion7,12,26. We sought to compare the biomechanical behavior of these reciprocal bands to see how they resist valgus moments and to determine if they differ in their susceptibility to injury depending on the angle of flexion of the elbow.
In a pilot study, ten fresh-frozen human cadaveric elbows were dissected to establish the anatomy of the medial collateral ligament. The ulnar nerve was released from the cubital tunnel and was reflected anteriorly. The common flexor tendons were detached near their humeral insertion and were reflected distally. The ligaments were inspected visually while the joint was flexed and extended. The lateral structures were then released, and the joint was opened to allow inspection of the medial collateral ligament from inside the medial aspect of the joint. Randomly selected specimens were frozen and sectioned for histological examination of the ligament insertions. An additional eighteen fresh-frozen human cadaveric elbows were tested biomechanically. The medial collateral ligament was exposed by transposition of the ulnar nerve. In every elbow, the anterior bundle had easily distinguishable anterior and posterior bands that tightened in reciprocal fashion (Fig. 1). Before biomechanical testing, the isometric fibers between the anterior and posterior bands were marked with a 5-0 nylon suture. During testing, the anterior band, the posterior band, and the posterior bundle were cut in four sequences.
The forearm was transected twenty centimeters distal to the joint line, and the distal ends of the radius and ulna were embedded in epoxy to fix the forearm in neutral pronation-supination. Soft tissues, including the proximal half of the interosseous membrane, were left intact. The humerus was transected twenty centimeters proximal to the joint line and was potted in epoxy at an angle such that the transepicondylar axis was horizontal when the joint was mounted in a specially designed testing apparatus, which was attached to a load-controlled materials-testing machine (MTS, Minneapolis, Minnesota)23. The axis of varus-valgus rotation of the elbow was perpendicular to the flexion-extension axis of the humerus and to the long axis of the ulna (Fig. 2). This definition of varus-valgus rotation has been used in previous studies9,18,26. Only the angle of flexion of the elbow was controlled; the other five degrees of freedom, including axial rotation of the forearm, were unconstrained.
Two-newton-meter varus and valgus torques were applied for three seconds, and torque, varus-valgus rotation, and rotation of the forearm were recorded with use of a personal computer. Two newton-meters of torque was chosen on the basis of a pilot experiment in which torques of more than 2.5 newton-meters resulted in avulsion of the ligament. This amount of torque is thought to be similar to that applied to the elbow during physical examination but is much less than that encountered during hard throwing2,24. The elbows were randomly assigned to one of four cut sequences (Table I). All cuts were made with a number-11 scalpel blade without removal of the specimen from the testing apparatus. After each structure was cut, the elbows were tested in 30, 60, 90, and 120 degrees of flexion. Each test was performed twice during the sequence, and the torque-rotation curves were highly reproducible.
The stiffness of the testing apparatus was determined by application of valgus torques to a potted steel bar and subtraction of the angular displacements from the experimental results. Tests were performed with and without a twenty-two-newton joint load applied along the axis of the ulna. Joint-loading diminished varus-valgus rotations without affecting trends; only the results of the tests that were performed without joint-loading are reported here. Valgus rotation of the elbow was accompanied by coupled internal rotation of the forearm relative to the humerus, which was also measured. The torque-rotation curves were plotted and were compared qualitatively to ascertain trends. The mean amount of varus-valgus rotation at two newton-meters of varus torque, zero newton-meters of torque, and two newton-meters of valgus torque was determined for each angle of flexion and each cut. The means were compared with analysis of variance, and correlations were tested with Pearson and Spearman correlation coefficients. The level of significance was p < 0.05 with use of a two-tailed test. The three tested components of the medial collateral ligament (the anterior and posterior bands and the posterior bundle) were ranked as primary, secondary, or tertiary restraints according to the test results. A structure was considered to be a primary restraint if cutting it caused increased valgus rotation in an otherwise intact elbow. Two structures were considered to be co-primary restraints if cutting either one by itself did not increase rotation but cutting both did. A structure was considered to be a secondary restraint if cutting it after the primary structure had been sectioned caused an additional increase in valgus rotation. Similarly, a structure was considered to be a tertiary restraint if cutting it after the primary and secondary structures had been sectioned caused an additional increase in valgus rotation18.
Anatomical
Biomechanical
Results at Each Angle of Flexion At 60 degrees of flexion, only the anterior band was sufficient to prevent increased valgus rotation when the other structures were cut. When the other two structures were intact, cutting of the anterior band significantly increased valgus rotation by a mean of 1.7 degrees (p < 0.001) (Fig. 3). Cutting of the posterior band or the posterior bundle after the anterior band had been sectioned did not cause an additional increase in valgus rotation. Thus, at 60 degrees of flexion, the anterior band was the primary restraint to valgus rotation and neither the posterior band nor the posterior bundle was a secondary restraint. At 90 degrees of flexion, only the anterior band was sufficient to prevent abnormal valgus rotation when the other structures were cut. When the other two structures were intact, cutting of the anterior band significantly increased valgus rotation by a mean of 1.0 degree (p = 0.03) (Fig. 3). Cutting of the posterior band after the anterior band had been sectioned caused an additional, significant increase in valgus rotation of 2.3 degrees (p = 0.02). In contrast, cutting of the posterior bundle after the anterior band had been sectioned did not cause an additional increase in valgus rotation. The anterior band was the primary restraint and the posterior band was a secondary restraint to valgus rotation at 90 degrees of flexion. At 120 degrees of flexion, cutting of either the anterior band (Fig. 3) or the posterior band (Fig. 4) alone did not result in abnormal valgus rotation; however, cutting of both bands (Fig. 5) significantly increased valgus rotation (p = 0.01). Cutting of the anterior band after the posterior band had been sectioned caused a mean increase in valgus rotation of 3.4 degrees (p = 0.019), and cutting of the posterior band after the anterior band had been sectioned caused a mean increase of 2.3 degrees (p = 0.07). Thus, the anterior and posterior bands of the anterior bundle were co-primary restraints at 120 degrees of flexion. The effect of the angle of flexion on valgus rotation after each structure had been sectioned was evaluated with least-squares linear-regression analysis. When all structures except the posterior band were cut, a negative correlation was found between the angle of flexion and the amount of valgus rotation (p = 0.03 and r = 0.54, Pearson correlation coefficient) (Fig. 4). No relationship was found between the angle of flexion and the amount of valgus rotation for the other structures or combinations of structures (Figs. 3 and 5).
Complete Sectioning of the Anterior Bundle
Complete Sectioning of the Medial Collateral Ligament
We found that the anterior bundle arose from the inferiormost aspect of the medial epicondyle and not from the anterior surface. This finding confirmed those of Fuss7, An and Morrey1,17, and O'Driscoll et al.22 but was not consistent with the anatomical findings that have been described in other reports6,11,15,27. We also found that the anterior bundle inserted immediately adjacent to the joint surface on the ulna near the sublimis tubercle, as was reported by Timmerman and Andrews29. On visual inspection, the anterior bundle was composed of two bands that appeared to tighten in reciprocal fashion as the elbow was flexed and extended. The posterior bundle was thickened joint capsule that tightened and fanned out to form a sharp edge with the elbow in flexion. The posterior bundle formed the floor of the cubital tunnel and appeared to compress the ulnar nerve as the elbow was flexed21. The ulnar nerve crossed the anterior bundle near its insertion on the sublimis tubercle.
Anterior and Posterior Bands We found that the anterior band was the primary restraint to valgus rotation at 30, 60, and 90 degrees of flexion; that the anterior and posterior bands were co-primary restraints at 120 degrees of flexion; and that the posterior band was a secondary restraint at 30 and 90 degrees of flexion. The posterior bundle did not appear to resist valgus rotation unless all other structures were cut, except at 30 degrees of flexion where it was considered to be a secondary restraint along with the posterior band. It was interesting that the anterior band remained a primary or co-primary restraint even at greater degrees of flexion, despite the visual appearance of decreasing tension. This finding suggests that some of the fibers of the anterior band were nearly isometric, which is contrary to the concept of reciprocal tightening of the anterior and posterior bands. An alternative explanation of the dominance of the anterior band is the mechanical advantage that it gains from being more parallel to the ulna than the posterior band is. Another possibility is that the anterior band is dominant because it is stiffer than the posterior band; however, the relative stiffness of these structures was not quantitated in the present study. Several other experimental findings support the concept of reciprocal tightening of the anterior and posterior bands. The dominance of the anterior band decreased with increasing flexion, and the posterior band tightened with flexion. Close visual inspection of the twenty-eight cadaveric elbows in the present study demonstrated consistent reciprocal tightening of the bands. These data suggest that the anterior and posterior bands have distinct biomechanical roles, that these structures may be injured separately, and that the type of injury may depend on the angle of flexion of the elbow at the time of injury. Isolated injury of the anterior band may occur when the elbow is between 90 degrees of flexion and full extension, when the anterior band is the primary restraint to valgus rotation. Combined injury of the anterior and posterior bands may occur at greater degrees of flexion, when the bands are co-primary restraints. Injury of the posterior bundle is unlikely unless the anterior bundle is completely disrupted, as in gross subluxation or dislocation of the joint. The valgus torque applied in this experiment was two newton-meters, whereas a throwing athlete may generate much higher torque about the elbow24. Thus, the small increases in valgus rotation caused by sequential sectioning of the structures in the present study may represent clinically important instability in the throwing athlete.
Physical Examination of the Medial Collateral Ligament The results of the present study suggest that clinical testing for complete tears of the anterior bundle should be performed with the elbow in 90 degrees of flexion. The difference in valgus rotation between the intact elbows and those in which the anterior band had been sectioned was significant only at 60 and 90 degrees of flexion. Also, the difference in valgus rotation between the intact elbows and those in which the entire anterior bundle had been sectioned was significantly greater at 90 degrees than at 30 degrees of flexion. Mean increases in valgus rotation caused by partial or complete sectioning of the anterior bundle ranged from 0 to 3.2 degrees. Although cuts made with a scalpel blade may not duplicate the effect of interstitial tears of the ligament, these findings suggest that tears of the anterior bundle may be undetectable with standard manual examination.
Operative Reconstruction of the Medial Collateral Ligament Partial tears of the anterior bundle may not produce abnormal valgus rotation on manual stress tests. This may account for the failure of non-operative treatment in some throwing athletes who have normal findings on physical examination. The optimum amount of flexion in which the elbow should be immobilized after an anatomical repair of the anterior bundle has not been defined. Morrey and An observed slight lengthening of the anterior bundle with the elbow in flexion17. Our data support the concept of reciprocal tightening of the anterior and posterior bands, which suggests that the anterior bundle is isometric when considered as a whole. Immobilization with the elbow in some degree of flexion may minimize stress on the graft in the critically important anterior band of the anterior bundle.
Comparison with Previous Biomechanical Studies An et al.3 and Morrey et al.18 examined the medial collateral ligament of the elbow with a Polhemus Isotrak device (McDonnell Douglas Electronics, Colchester, Vermont), which was used to measure flexion, valgus rotation, and axial rotation of the ulna throughout the range of motion. We found a similar degree of valgus rotation in the intact elbows but did not observe any decline in valgus rotation with increasing flexion. This difference may be related to the experimental design. The study by Morrey et al. is not useful for comparison of the components of the medial collateral ligament because only one elbow was tested with sequential sectioning of the structures. Søjbjerg et al. tested the medial collateral ligament by sequential sectioning in cadaveric specimens and found that cutting of the entire anterior bundle increased valgus rotation by a maximum of 11.8 degrees with the elbow at 70 degrees of flexion27,28. In contrast, we found that cutting of the anterior bundle increased valgus rotation by only 3.2 degrees. Also, we found that subsequent sectioning of the posterior bundle allowed subluxation of the joint, whereas Søjbjerg et al. found no additional increase in valgus rotation. We confirmed the finding of Søjbjerg et al. that primary sectioning of the posterior bundle does not affect valgus rotation.
Limitations of the Experimental Technique We were unable to test the specimens in full extension or full flexion. The maximum extension and flexion of the specimens varied slightly. The testing apparatus did not allow fully variable positioning, as this would not permit rigid clamping of the pot in the materials-testing machine. Therefore, 30, 60, 90, and 120 degrees were chosen as the angles of flexion for testing. Results derived with the load-controlled technique may change if the sequence of cuts is changed. This may be true if the testing method is destructive to the specimens. In the present experiment, however, we varied the sequence of the cuts and found that the angular displacement for a combination of cuts did not depend on the order in which the cuts were made. Also, we believe that the load-controlled technique was analogous to the physical examination, during which the examiner applies a standard torque at a standard rate and measures the resulting angular displacement.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Technorati What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |