Biomechanical Consequences of Coracoclavicular Reconstruction Techniques on Clavicle Strength

OBJECTIVES: Lateral clavicle fractures have been reported following coracoclavicular (CC) ligament reconstruction with bone tunnels through the clavicle. Several techniques for CC reconstruction with different drill-hole measurements have been described. Our objective was to evaluate clavicle weakening related to tunnel diameters for common CC-reconstruction techniques. METHODS: Testing was performed on 2 groups of 18 matched pair clavicles, which were randomly distributed between groups. There were no significant differences between the groups regarding bone mess density (BMD), clavicle width, age, and gender. One clavicle from each pair was prepared according to one of two reconstruction techniques; the contralateral clavicle was left intact. Both techniques placed 2 tunnels through the medial clavicle, 30 mm and 45 mm from the lateral border. Group 1 (mean age: 53, range: 44-63; mean BMD: 0.48, range: 0.39-0.59) was prepared with 2.4 mm tunnels and augmentation devices. Group 2 (mean age: 56, range: 45-63; mean BMD: 0.47, range 0.35-0.61) was prepared with 6.0 mm tunnels with hamstring grafts and tenodesis screws. A 3-point bending load was applied to the distal clavicles at 15 mm/min until failure. Ultimate failure load and anterior-posterior width 45 mm medial from the lateral border were recorded for each specimen. Strength reduction was determined as the percent reduction in ultimate failure load between paired intact and surgically prepared clavicles. Relative tunnel size was determined as the quotient of tunnel diameter and clavicle width. An independent observer performed all clavicle width measurements. Non-parametric statistics were used (MWU, Kendall’s Tau). RESULTS: The 6.0 mm technique significantly reduced clavicle strength relative to intact (p = 0.02) and caused significantly more strength reduction than the 2.4 mm technique (p = 0.02) (Figure). The 2.4 mm technique was not significantly different from intact. All but one fractures occurred at the medial tunnel. Clavicle width at the medial hole varied highly (mean: 18.1 mm, range: 12.3 - 27.1 mm). There was a significant approximately linear correlation between clavicle width and strength reduction (p = 0.04, tau = -0.36) and between relative tunnel size and strength reduction (p < 0.01, tau = 0.51). Therefore, clavicle strength reductions of 30% and 50% relative to the intact state can be expected with relative tunnel diameters of 34.5% and 49.8% of the clavicle width, respectively. The intra-observer correlation coefficient of the width measurement was excellent (0.99; 95% CI: 0.98 - 0.99). CONCLUSION: Coracoclavicular ligamentous reconstruction with 6.0 mm tunnels, graft, and tenodesis screws caused significantly greater decreases in the strength of the clavicle when compared to 2.4 mm tunnels with augementation devices and undrilled controls. Additionally, strength reductions correlated highly with the ratio of tunnel width relative to overall clavicle width. This information can help optimize techniques for reconstructing unstable distal clavicles and can influence the intraoperative decision-making process based on the individual clavicle width.