An in Vitro Biomechanical Model of Differing Pedicle Screw Configurations for Long Construct Segmental Thoracic Fixation.

Background: The optimum pattern of pedicle screw (PS) fixation during long-segment thoracic fixation has not been determined.

Objective: To evaluate rod stress and construct stability with minimal, alternating, skipped, and bilateral PS constructs in the iatrogenically destabilized thoracic spine.

Methods: Eight cadaveric thoracic specimens (T3-T12) were initially tested intact to ±5 Nm using a custom 6 degree-of-freedom spine testing apparatus in flexion-extension (FE), lateral bending (LB), and axial rotation. Specimens were instrumented with T4-T10 bilateral PS, with Ponte osteotomies to introduce instability. Rods were bent to fit the PS and then spines were tested with the minimal, alternating, skipped, and bilateral fixation patterns. Range of motion (ROM) was calculated from T4-T10 and segmentally. In addition, strain gauges fixed to the spinal rods measured rod stress under FE and LB. Results were compared using ANOVA and post hoc Holm Sidak tests.

Results: All fixation patterns provided significant reductions in ROM with respect to the intact spine. In all motion planes, minimal provided the least amount of rigidity, while bilateral provide the greatest; however, no statistically significant differences were detected in FE. In LB and axial rotation, skipped, alternating, and bilateral were all significantly more rigid than minimal (P < .01). Rod strains were greatest under LB and correlated with overall construct ROM, where bilateral had significantly lower strain than the other patterns (P < .05).

Conclusion: All constructs effectively decreased thoracic ROM. There was significant improvement in stabilization and decreased rod stress when more fixation points beyond the minimal construct were included.