Biomechanical comparison of lumbosacral fixation using Luque-Galveston and Colorado II sacropelvic fixation: advantage of using locked proximal fixation

Abstract

Study design: Biomechanical evaluation of sacropelvic fixation strategies as they apply to neuromuscular scoliosis.

Objectives: The primary objective was to compare the rigidity of 2 methods of sacropelvic fixation (Galveston vs. Colorado II). The secondary objective was to evaluate the effect on construct rigidity by adding a pair of L1 pedicle screws to a Luque wire construct.

Summary of background data: The Galveston modification to the Luque rodding system has become standard for treating childhood and adolescent neuromuscular scoliosis. The Galveston method provides reasonable lumbo-pelvic fixation with a relatively simple method of insertion. Clinical reviews of sagittal plane stability in neuromuscular patients with Galveston fixations performed at our institution have led to concerns regarding the technique's ability to maintain proper lumbar lordosis. This concern has generated interest in evaluating biomechanical stability of more rigid fixation methods in these long spino-pelvic constructs. As such, the following biomechanical study evaluated lumbosacral stability of 2 sacropelvic fixation methods: the standard Luque-Galveston method and the Colorado II sacropelvic fixation method using the Chopin plate-screw block. As a secondary interest, evaluations of the rigidity of the proximal construct when using pedicle screw fixation were completed. It was hypothesized that one additional point of rigid fixation at the thoracolumbar junction may make substantial improvement in rigidity to the otherwise Luque construct.

Methods: Lumbo-pelvic segments of human cadaveric specimens were instrumented with L1 pedicle screws, sublaminar wires between L2 and L5, and sacropelvic fixation with either Galveston rods or Colorado II sacropelvic plates using S1 screws, S2 alar screws, and iliac screws. Tests were conducted for physiologic flexion-extension and torsional loading. Construct stiffness between L1-S1 was determined for each specimen. Motion measurement data were collected between L1-L5 and L5-S1 using a noncontact marker system. Statistical analysis included a 2 -way analysis of variance (dependent variables: construct/locked screw) with the Tukey post hoc correction for multiple comparisons (P < 0.05).

Results: The flexion and extension bending stiffness of the construct was similar between the Galveston and Colorado II constructs. Both constructs were stiffer when the L1 screws were locked rigidly to the rod. Torsional stiffness followed similar trends with no significant difference between the systems, although more rigid in more cases when the L1 screws were locked to the rod. Regarding limiting L5-S1 motion during flexion and extension loading, the Colorado II construct did so to a higher degree. There was no difference in torsional motion between the 2 constructs. Locking the L1 pedicle screws reduced torsional motion but had no effect on flexion-extension motion at L5-S1.

Conclusion: The 2 methods of sacropelvic fixation provided similar construct stiffness, although the Colorado II method had less L5-S1 motion on flexion-extension testing, and the Galveston construct tended (although not statistically) to be stiffer in torsional loading. The addition of a pair of L1 pedicle screws increased the construct stiffness for both constructs by approximately 50%.