Biomechanical evaluation of short-segment posterior instrumentation with and without crosslinks in a human cadaveric unstable thoracolumbar burst fracture model

Abstract

Study design: This study evaluates the biomechanical characteristics of spinal instrumentation constructs in a human unstable thoracolumbar burst fracture model simulated by corpectomy.

Objective: To compare the biomechanical characteristics of short-segment posterior instrumentation, with and without crosslinks, in a human unstable burst fracture model simulated by corpectomy.

Summary of background data: Unstable thoracolumbar burst fractures are serious injuries, and their management remains controversial. Some authors advocate the use of short-segment posterior instrumentation for certain burst fractures. Whether crosslinks contribute additional stability has not been determined.

Methods: Six fresh frozen human spines (T10-L2) were potted to isolate the T11-L1 segments, and biomechanically tested in axial rotation, lateral bending, flexion, and extension. A custom spine testing system was used that allows motion with 6 degrees of freedom. After testing was completed on intact specimens, a corpectomy was performed at T12 to simulate an unstable burst fracture with loss of anterior and middle column support. Short-segment transpedicular instrumentation was then performed from T11 to L1. Each specimen was retested with 1, 2, or no crosslinks. Construct stiffness and motion data were analyzed with each intact specimen serving as its own internal control.

Results: Torsional stiffness in axial rotation was significantly increased (P < 0.05) in short-segment fixation constructs with 1 and 2 crosslinks, but none was restored to the preinjury baseline level. Significant reductions in standardized motion were also achieved with 1 and 2 crosslinks compared to no crosslinks (P < 0.05), but they remained greater than baseline. Crosslinks significantly increased stiffness and decreased motion in lateral bending, beyond the baseline level (P < 0.05). In flexion, all constructs had significantly decreased stiffness and increased motion compared to the intact specimen (P < 0.05), with crosslinks providing no additional benefit. Conversely, none of the constructs demonstrated a significant change in extension compared to baseline (P > 0.05). When attempting to load the constructs to failure, screw pullout was seen in all specimens.

Conclusion: Crosslinks, when added to short-segment posterior fixation, improve stiffness and decrease motion in axial rotation, but do not restore baseline stability in this corpectomy model. Short-segment posterior fixation is also inadequate in restoring stability in flexion with injuries of this severity. Short-segment posterior instrumentation alone can achieve baseline stability in lateral bending, and crosslinks provide even greater stiffness.