3D rod shape changes in adolescent idiopathic scoliosis instrumentation: how much does it impact correction?

Flattening of rods is known to reduce the correction capability of the instrumentation, but has not been studied in 3D. The aim is to evaluate the rods shape 3D changes during and immediately after instrumentation, and its effect on 3D correction.

The 5.5 mm CoCr rods of 35 right thoracic adolescent idiopathic scoliosis patients were measured from rod tracings prior to insertion, and reconstructed in 3D from bi-planar radiographs taken intra-operatively after the correction maneuvers and 1 week post-operatively. The rod bending curvature, maximal deflection and orientation of the rod's plane of maximum curvature (RPMC) were computed at each stage. The relation between rod contour, kyphosis and apical vertebral rotation (AVR) was assessed.

Main thoracic Cobb angle was corrected from 58° ± 10° to 15° ± 8°. Prior to insertion, rods were more bent on the concave side (curvature/deflection: 39° ± 8°/25 ± 6 mm) than the convex side (26° ± 5°/17 ± 3 mm). Only the concave rod shape changed after the correction maneuvers execution (flattening of 21° ± 9°/13 ± 7 mm; p < 0.001) and stayed unchanged post-operatively. After instrumentation, the RPMC was deviated from the sagittal plane (concave side: 27° ± 19°/convex side: 15° ± 12°). There was a significant association between kyphosis change and the relative concave rod to spine contour (rod curvature-pre-operative kyphosis) (R ² = 0.58) and between AVR correction and initial differential concave/convex rods deflection (R ² = 0.28).

Correction maneuvers induce a significant change of the concave rod profile. Both rods end in a plane deviated from the sagittal plane which is representative of the spinal curvature 3D orientation. Differential rod contouring technique has a significant impact on the resulting thoracic kyphosis and transverse plane correction.