Étude biomécanique d'instrumentations antérieures pour le traitement de la scoliose

Pre-instrumentation reduction, consisting to reduce scoliotic deformation before implants insertion, could be a mean to improve surgery correction in frontal and sagittal plane. The project's hypothesis is that pre-instrumentation reduction of scoliotic deformity has a clinically and statically significant influence (p<0.05) on the correction in frontal and sagittal plane of the scoliotic spine during an anterior instrumentation. The correction in frontal and sagittal plane is characterised by reduction of the Cobb angle for the principal curvature, the thoracic kyphosis, the lumbar lordosis, intervertebral disc wedging in the frontal plane and internal strain of the epiphyseal growth plate. The main objective of the project is to determine the contribution of this reduction in the overall changes seen post-operatively.

Finite elements model (FEM) of scoliotic spine was developed to simulate the different steps of the surgery: per-operative positioning of the patient, surgery manipulation and postoperative positioning of the patient (comeback in stand-up position). Implants modeled include a cable, linked to vertebras by rigid link elements. Different study configurations were modeled by changing the force applied by the surgeon to the spine before instrumentation and the cable characteristic: material type, initial tension, distance between the cable and vertebras.

Three different designs were modeled for 6 patients. For all designs a steel cable with a 50N initial tension was used with a 0.5 cm distance to the vertebras. Pre-instrumentation reduction was obtained by no applying external effort to the spine (design 1) or obtained by applying a 50N force laterally to the spine (design 2) or by applying a 150N force to the spine (design 3). Results of designs 1 and 2 and results of designs 1 and 3 were compared by a paired Student t-test. The same study design was used to evaluate impact of the variation of initial tension in the cable.

To study the influence of the pre-instrumentation reduction in relation to other surgery parameters, a mixed experimental plan 3²2 ¹ was done for 6 patients. Independent parameters was pre-instrumentation reduction obtained by a 50N to 150N force applied to the spine, material of cable (steel, Nitinol, polyethylene) and 0.5 cm to 1 cm distance between cable and vertebras (just for one patient). Dependant parameters were the same as the sensibility study. Finally, to compare the influence of the most important parameters determined during previous studies, a last experimental plan 2 ³ was done. Independent parameters were pre-instrumentation reduction obtained by a 50N to 150N force applied in the spine, material of cable (steel, polyethylene) and 50N to 150N initial tension in the cable. Dependant parameters were the same as previous experimental plans.

Maximal error in compression and lateral flexion test is 0.25 mm, in bending test 0.25 mm and 1.1° and in torsion test 0.3° compared with literature data. The sensibility study shows that only the annulus mechanical properties have a clinically and statically significantly influence on simulated correction. Mean error of the per-operative positioning of the patient in the model relative to radiographic measurements is 2.5° for Cobb angle, 2.3 mm for thoracic AVT, 1.4 mm for lumbar AVT and 3.6 mm for height. Mean error to the post-operative positioning of the patient in the model relative to radiographic measurements is 0.3° for Cobb angle, 0.05 mm for thoracic AVT, 0.05 mm for lumbar AVT and 0.05 mm for height.

Exploitation of the model demonstrates that pre-instrumentation reduction of the scoliotic deformation has a clinically and statistically significant influence on correction only in the frontal plane. This reduction could have clinically and statistically significant influence on the intervertebral discs wedging in the frontal plane and the strain's asymmetry correction in the growth plate. But the material type of cable has an even greater influence than the pre-instrumentation reduction. On the other hand, distance between cable and vertebras has a negligible influence on correction. Finally initial tension in the cable has significant influence on correction only for flexible cable. (Abstract shortened by UMI.)