Influence de la morphologie du rachis cervical sur les dommages de la moelle spinale dans le syndrome centromédullaire traumatique

Abstract

The central cord syndrome (CCS) is a subset of incomplete spinal cord injuries (SCI-I) with “disproportionately more motor impairment of the upper than of the lower extremities”. It is diagnosed in 30 to 40 % of SCI-I cases and is more prevalent in the elderly population as a consequence of a forward fall. However, conflicting results are provided by the literature on this syndrome’s genesis, its management, and its prognostics. It has been proposed that the main cause of result discrepancies is the lack of an objective CCS definition, leading institutions to diagnose this syndrome based on different thresholds of expressivity. Considering that the vertebral morphology is influenced by demographic factors, could the spine morphology influence damage sustained by the spinal cord after a traumatic event commonly associated with a CCS diagnostic, and could it contribute to the variability of results and symptoms provided by the literature? Firstly, a retrospective analysis of a prospective clinical database has been performed on 336 patients hospitalized for a cervical SCI-I. The CCS expressivity has been binarized into two classes based on the cohort’s median expressivity value (0.22), and a logistic regression has been used to investigate the influence of seven demographic and clinical factors. Females are more likely than males to present higher expressivity after a fall, and are less able than men to present high expressivity over the full range of SCI-I severity. A nnU-Net segmentation model has then been trained on the specific task of cervical vertebrae segmentation in CT-Scan images, and an algorithm to automate anatomical measurements of segmented vertebrae has been developed and used on a cohort of 450 adults. An ascending hierarchical clustering on principal components has then been performed on these measurements, and it exhibited that the human morphology can be summarized by four cervical spine archetypes, that the global size of vertebrae is linked to the individual’s sex but is not correlated to the spinal canal size and neither to the intervertebral disc thickness. Finally, 21 hyperextension simulations of the C3-C4 functional spinal unit, based on the aforementioned archetypes, exposed that the vertebral size (i.e. the individual’s sex) influences the spine’s stiffness and the extent of damage sustained by the spinal cord. Damages were consistent with CCS genesis and previous results from this study: females with a stenotic spinal canal would sustain too much axonal in order to present a high expressivity profile, and males who fell forward would sustain too few axonal damage. More simulations are required, along with implementation of degenerated tissues material properties. Nevertheless, the spinal morphology appears to influence the CCS genesis and expressivity as soon as the primary injury. By combining a clinical study and a fundamental numerical study, it has been demonstrated that individuals usually excluded from CCS studies, paradoxically, are the elderly if they are not close to a 50th-centile in morphology. This study therefore supports the call to standardize an objective definition of the CCS and rekindle the debate on the pertinence of cervical stenosis prophylactic surgeries.

Date	7 Aug 2026
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Yvan Petit (Supervisor) & Jean Marc Mac-Thiong (Co-supervisor)