Observation of sub-clinical neurological abnormalities has led to the proposal of a neuro-developmental etiologic model for AIS. Our research group have demonstrated longer latency in somatosensory–evoked potential (SSEP) and impaired balance control in AIS subjects. A previous pilot study compared the regional brain volume between right thoracic AIS subjects and normal controls. Significant regional brain differences were found relating to corpus callosum, premotor cortex, proprioceptive and visual centers. Most of these regions involved the brain unilaterally, indicating there might be abnormal asymmetrical development in the brain in right thoracic AIS. In this pilot study, we investigated whether similar changes are present in left thoracic AIS patients who differ from matched control subjects. Nine AIS female patients with atypical left thoracic AIS (mean age 14.8, mean Cobb angle 19°) and 11 matched controls as well as 20 right thoracic AIS (mean Cobb angle 33.8°) and 17 matched controls, underwent three-dimensional isotropic magnetization prepared rapid acquisition gradient echo (3D_MPRAGE) magnetic resonance (MR) imaging of the brain. Fully automatic morphometric analysis was used to analyse the MR images; it included brain-tissue classification into grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF). and non-linear registration to a template brain. Tissue densities were compared between AIS subjects and controls. There was no significant difference between AIS subjects and normal controls when comparing absolute and relative (i.e. brain-size adjusted) volumes of grey and white matter. Using voxel-based morphometry, significant group differences (controls > left AIS) were found in the density of WM in the genu of the corpus callosum, the left internal capsule (anterior arm) and WM underlying the orbitofrontal cortex of the left hemisphere. The above differences were not observed in the right AIS group. This first controlled study of regional tissue density showed that corpus callosum, which is the major commissural fiber tract, was different in the atypical left thoracic scoliosis while significant regional brain changes have not yet been found in those with typical right thoracic scoliosis. Further investigation is warranted to see whether the above discrepancy is related to laterality of the scoliotic curves and infratentorial neuroanatomical abnormalities. A larger sample and a longitudinal study is required to establish whether the brain abnormalities are predictive of curve progression.

Introduction: Observation of sub-clinical neurological abnormalities has led to the proposal of a neuro-developmental etiologic model for adolescent idiopathic scoliosis (AIS). We have previously demonstrated prolonged latency in somatosensory evoked potentials (SSEP) and impaired balance control in subjects with AIS. Furthermore we have compared regional brain volumes in right thoracic AIS subjects and normal controls. Significant neuro-anatomic regional differences were observed in the corpus callosum, premotor cortex, proprioceptive and visual centers of the AIS subjects compared to control subjects. Most of these regional differences involved the brain unilaterally, indicating there may be abnormal asymmetrical development in the brain of subjects with right thoracic AIS.

Methods: Following ethical committee approval a total of 29 subjects with AIS were recruited. Patients with congenital, neuromuscular or syndromic scoliosis were excluded from the study. Twenty-eight age- and sex-matched controls were recruited from local schools. All recruits underwent three-dimensional isotropic magnetization prepared rapid acquisition gradient echo (3D_MPRAGE) magnetic resonance (MR) imaging of the brain. Modern morphometric analyses of the MR images were carried out including classification of tissue into grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF). Tissue densities were compared between AIS subjects and controls. Comparisons were made between those subjects with left thoracic AIS (n=9) and age and sex-matched controls (n=11) and those subjects with right thoracic AIS (n=20) and age and sex-matched controls (n=17).

Results: For subjects with left thoracic curves the mean Cobb angle was 19 degrees. For subjects with right thoracic curves the mean Cobb angle was 33.8 degrees There was no significant differences observed between AIS subjects and normal controls when comparing both absolute and relative (i.e. adjusted for brain size) volumes of GM and WM. However voxel-based morphometric analysis identified significant differences in the density of WM in the genu of the corpus callosum, the left internal capsule and WM underlying the left orbitofrontal cortex when comparing those subjects with left thoracic scoliosis to controls. The above differences were not not observed when those subjects with right thoracic scoliosis were compared to controls..

Discussion: This controlled study of regional brain tissue density has demonstrated important differences in the corpus callosum, the left internal capsule and the left orbitofrontal cortex when the brain of those subjects with left thoracic scoliosis is compared to age and sex matched controls. In this study significant regional brain differences have not been identified in those subjects with right thoracic scoliosis. Further studies are warranted to ascertain whether these morphologial differences in the brain are linked with the etiopathogenisis of left sided thoracic scoliosis. A larger sample and a longitudinal study are required to establish whether brain abnormalities are predictive of curve progression.