Étude biomécanique de la spondylolyse et du spondylolisthésis chez l'enfant: Étude de cas

Spondylolisthesis is a postero-anterior slippage of a vertebra on the one directly below. It mainly occurs after spondylolysis, which is a stress fracture of the pars of the cranial vertebra, or after a high dysplasia of its posterior bony elements. Spondylolisthesis affects about 6% of the general American population and usually occurs at the lumbosacral junction. The literature describes classification systems for low- and high-grade spondylolisthesis. For low-grade spondylolisthesis (grade 1 and 2), there are the shear-type patients and the nutcracker-type patients, and the pelvic incidence (PI) and sacral slope (SS) are the main measurement parameters. For high-grade spondylolisthesis, the SS and pelvic tilt (PT) are the main measurement parameters, and there are the "balanced pelvis"-type patients and the "retroverted pelvis"-type patients.

Forces responsible for the progression of both pathologies are poorly documented in literature. There are only few studies of pediatric spondylolisthesis. The aim of this project is to develop a detailed and personalized finite element model (FEM) to investigate the biomechanics of pediatric L5-S1 spondylolysis and spondylolisthesis. This model can be personalized for patients with low- or high-grade spondylolytic spondylolisthesis. A detailed analysis can be done with the FEM in the epiphyseal growth plates and osseous endplates in the lumbosacral junction. Patients were selected for this study according to the classification systems reported in the literature.

The geometry of the spine, pelvis and rib cage was reconstructed in 3D using biplanar radiographs of low- and high-grade spondylolisthesis patients. Taking into consideration the effects of muscles and relevant inter and para-vertebral connective tissues, a personalized biomechanical model was established with enhanced details of the L4 to pelvis segment. Seventeen anatomical landmarks were used for the 3D reconstruction of thoracic and lumbar vertebrae and 23 were used for the pelvis. The 3D coordinates of the nodes of the existing bony FEM were then deformed using dual kriging to fit these reconstruction points. The FEM fits all the landmarks identified by the reconstruction process, and the 3D coordinates of other nodes are statistically determined. Mechanical properties of anatomical structures were found in the literature.

Four different patients have been selected for this study according to the classification systems reported in the literature. Different loading and physiological movements were studied. A flexion of 30° and an extension of 40° were simulated numerically. A bilateral lysis was created by manually removing posterior elements of L5. The combined effects of muscles and gravity were based on a gradient and opportunistic coordinate search optimisation process. This fast method consists in the evaluation of different start points of optimization, equally spaced, in order to identify which is the best. For a chosen start point, the gradient method finds the muscles' contribution necessary to execute the movement. Different start points are tested and when the best point is found among those selected, the interval between separated start points is divided in half. Optimization then proceeds from this initial point until the resulting geometry is the same as the 3D reconstruction. A fifth pathological patient that progressed from a grade 2 to a grade 3 spondylolisthesis has been used for the assessment of spondylolisthesis progression.

Principal and shear stress were calculated in the lumbosacral joint. Stress was studied more specifically at the epiphyseal growth plates and the osseous endplate of L5. For the fifth patient, stress was higher in grade 3 than in grade 2 spondylolisthesis. Compression stress on L5 was higher than that reported in some literatures for the nutcracker-type patient, whereas normal and shear stresses were higher in the shear-type patient than those reported on the nutcracker-type patient. Furthermore, for the shear-type patient, flexion movements induced high shear stress at the osseous endplate of L5, at the junction of its growth plate. Stresses were higher in high-grade spondylolisthesis than in low-grade spondylolisthesis.

Results suggest that an extension mechanism fracture of the pars is dominant for low-grade spondylolisthesis patients. The relative displacement between vertebrae is restrained by the facet joints which induce normal stress in the pars. Existing stress at the lumbosacral junction for the shear- and nutcracker-type patient could lead to a physis stress fracture of the vertebral body and then, further slippage at the growth plate. For both low-grade patients, compression on the anterior part of the S1 growth plate could be associated with its dome-shaped morphology change during the pathology's progression. Results from high-grade simulations however, suggest a higher grade of slippage increases the risk of further slippage over time.