Biomechanical contribution of transverse connectors to segmental stability following long segment instrumentation with thoracic pedicle screws.

STUDY
DESIGN: An in vitro biomechanical cadaver study of long segment thoracic pedicle screw constructs with transverse connectors (TC).
OBJECTIVE: To determine the resultant degree of motion of the instrumented thoracic spine after segmental pedicle screw instrumentation with and without TC. SUMMARY OF BACKGROUND DATA.: TC are generally not thought to be necessary with thoracic pedicle screw constructs, yet to date no study has reported the effect of TCs after all pedicle screw long thoracic fusions.
METHODS: Eight human cadaveric spines were potted and then instrumented from T4-T10 with bilateral 5.5 mm multiaxial titanium (Ti) pedicle screws and 5.5 mm contoured Ti rods. Specimens were tested with a six-degree-of-freedom spine stimulator in the intact condition, after instrumentation, after placement of 1 TC (3 different locations) and after placement of both TCs. Data were analyzed by loading modality (axial rotation, flexion-extension, and lateral bending) using one-way analysis of variance with an alpha of 0.05. Paired t tests were used for post hoc analysis with correction for multiple comparisons.
RESULTS: There was no difference with the addition of 1 or 2 TCs in terms of flexion-extension or lateral bending when compared to the instrumented condition (P > 0.05). Biomechanical testing of the long-segment thoracic constructs in axial rotation (torsion) loading modes generated the most significant findings of this study. After instrumentation with thoracic pedicle screws, T4-T10 full ROM was significantly reduced from the intact condition (P < 0.05). On average, TPS alone resulted in a 65% decrease in ROM. However, the addition of a transverse connector at 1 of the 3 positions tested yielded another 20% improvement in axial segmental stability as represented by further ROM reduction. These differences were significant from the TPS only group (no TCs), regardless of the TC position (P < 0.05). Furthermore, 2 TCs placed at the proximal and distal ends of the construct provided the greatest biomechanical axial stability to the instrumented specimens (P < 0.05). This was highlighted by an average of 35% ROM reduction from the stability level achieved with the TPS only constructs (P < 0.05), or an additional 15% improvement in axial stability over a single TC.
CONCLUSION: For long thoracic pedicle screw constructs, the addition of 1 or 2 TCs significantly decreases construct axial rotation, which is the primary plane of motion for the thoracic spinal region. A single TC contributed to a significant reduction of T4-T10 ROM (an additional 20%) relative to TPS fixation alone (P < 0.05), while the location of the TC within the construct was irrelevant. A second TC had an additive effect (an additional 15% reduction) on axial stability. (P < 0.05) Flexion-extension and lateral bending are not affected. Single TC significantly improves axial rotation stability in long thoracic pedicle screw constructs. Two crosslinks, however, are better than one.