The present study analyzed the ratios of VBHa to VBHp in the thoracic and lumbar spines of patients with AIS and non-oscoliotic adolescents. The results showed that the ratios in patients with AIS were significantly higher than those in adolescents without scoliosis; in other words, the mean VBHa in patients with scoliosis was longer not only in the thoracic vertebrae, but also in the lumbar vertebrae. More than 100 years ago, Meyer 14 and MacLennan15 stated that almost all cases of scoliosis were due to an inequality in growth between the posterior and anterior spinal columns. Roaf et al. subsequently reported that the basic lesion in scoliosis was the relative lengthening of the anterior components of the spine compared with the posterior elements.5 Professor Cheng et al. later applied whole-spine magnetic resonance imaging (MRI) to re-investigate the height of both the anterior and posterior vertebral components in girls with AIS and in normal subjects. This study confirmed the results of previous anatomical studies, and supported the consensus view that patients with thoracic AIS exhibit relatively faster growth of the anterior and slower growth of the posterior elements of the thoracic vertebrae. 6 Vertebral torsion may also play an important role in the process of scoliosis resulting from RASO.16–18 In a study using three-dimensional MRI, Birchall D et al19 previously supported the theory that torsion was a fundamental part of the early pathomechanical process of AIS; they further hypothesized that there must be a consistent, long-term deforming force that results in a rotational torque being applied to the growing vertebral body, and it is likely that this force is fundamental to the pathogenesis of the scoliotic curve. Although the primacy of the sagittal plane in the pathogenesis of AIS has been questioned,9–12 the obligatory involvement of the sagittal plane of the spine in curve initiation is generally accepted. In recent years, research on vertebral body wedging in the sagittal plane supports the claim that scoliosis can be initiated through hypokyphosis;20 however, some studies have shown that there is no significant difference between sagittal wedge angles in mild and moderate scoliosis, and sagittal wedging has not been found to be an important component of the local vertebral deformity. Therefore, we cannot support the theory that the precipitating factor for scoliosis initiation is anterior overgrowth or posterior tethering, which causes the spine to buckle. 21 However, only a limited number of studies have focused on the height of the lumbar vertebrae. In the present study, we found that the VBHa on the lumbar vertebrae grew faster, regardless of the scoliotic curve of the thoracic, thoracolumbar, or lumbar spine. The results showed that RASO arose simultaneously on the thoracic and lumbar vertebrae in AIS patients, regardless of where the scoliotic curve was, which is to say that the RASO was generalized not only on the vertebrae around the apex. However, Schlösser demonstrated that anterior overgrowth is not a generalized growth disturbance of the spine, but is instead confined to the area around the apex of both the primary and secondary curves.12 This disparity in results is probably associated with the inclusion of patients with different scoliosis severities (mild and moderate scoliosis in our study and severe scoliosis in Schlösser’s study).
Although the ratios at the same level on the thoracic and lumbar vertebrae were significantly different between the AIS and non-scoliotic groups in the present study, consistent trends between the T6 and L5 were found for in both groups, with ascension from T7 to T10, descension from T10 to T12, and ascension again from L1 to L5. The differences in ratios at the same level on different vertebrae between patients with AIS and healthy adolescents were very similar; therefore, the two curves of the ratios were nearly parallel. The trend of alteration in ratios in our study was similar to that observed in a previous study.22 As such, we speculated that RASO is not only a generalized condition in AIS, but that the velocities of overgrowth on different vertebrae are almost equivalent and coordinated with the growth of VBHp over the same segments. In other words, the anterior column of the vertebral body at the apex does not grow faster than the anterior column of the distal vertebrae. This phenomenon makes every vertebral wedging in the sagittal plane in patients with AIS consistent with the condition in healthy adolescents, and indicates that the spine is integrated to adapt to local deformities, such as scoliosis, to maintain spinal and truncal function in patients with AIS to the greatest extent possible.
Spinal sagittal malalignment in AIS with different segmental scolioses has been reported in several prior studies. The results showed that thoracic kyphosis (TK) was significantly higher in the lumbar curves than in King I, King II, and King III curves. However, the difference in TK between the thoracolumbar curves and other groups was not significant. The degree of lumbar lordosis also tended to be higher for patients with a lumbar curve, although not significantly.23 In another study, 49% of the curves presented sagittal malalignment in mild thoracic AIS, whereas only 13% of the (thoraco) lumbar curves and 6% of the nonscoliosis adolescents were hypokyphotic.4 Labrom et al recently corroborated these findings in a longitudinal MRI study of AIS, finding that the changes in thoracic major coronal curve angle were positively correlated with increases in vertebral body wedging angles.24 However, Mak et al found that all AIS patients had a similar degree of thoracic kyphosis regardless of coronal curve type.25 Another study also postulated that sagittal deformity is a generalized consequence seen across all scoliotic curvatures.26 The results of the present study are largely in agreement with those of Mak et al. However, ee found no significant differences in the ratios of thoracic, thoracolumbar, and lumbar scoliosis. The ratios were similar not only in the thoracic vertebrae, but also in the lumbar vertebrae despite the scoliotic site in the coronal plane. These results further confirm that RASO results from the overall effect on the spine rather than from the effect of local scoliotic levels in patients with AIS.
Previous studies have reported different results regarding the correlation between RASO and scoliosis severity. A significant positive correlation between the scoliosis severity score and the ratio of the differential growth between the anterior and posterior columns for each thoracic vertebra was found in a prior study conducted by Guo.6 In another study, the researchers also observed linear correlations between the coronal Cobb angle, axial rotation, and the anterior-posterior length difference (r = 0.729 for thoracic curves; r = 0.485 for (thoraco)lumbar curves).12 Vertebral wedging presented with mild scoliosis and increased as the scoliosis progressed.20 However, De Smet et al 27 found no correlation between scoliotic segmental kyphosis and the degree of frontal Cobb curve angle. In another study, Newell et al28 measured the changes in vertebral body height over time during scoliosis progression to assess how vertebral body height discrepancies change during growth, with the results showing that the degree of RASO was not related to the rate of progression or severity of the scoliotic curve, whereas AIS patients had a proportionally longer anterior column than non-scoliotic controls. The ratios of the different vertebrae did not correlate well with the coronal Cobb angles in this study. The observed significant correlations, which were very weak, were positive for T7(r = 0.190), T8(r = 0.195), and T11(r = 0.216), and negative for L5 (r= -0.186). These results suggest that anterior overgrowth was unaffected by scoliosis severity in patients with mild and moderate AIS. We speculated that the differences in the results were probably due to the recruited subjects having different scoliosis severities. In our study, we exclusively enrolled patients with severe scoliosis (Cobb angle > 45°), whereas patients with mild, moderate, and severe degrees were included in Guo’s study, and most subjects had serious curves in Schlösser’s study.
Our study had some limitations. First, this was a retrospective single-center study. Second, the ratios between T1-T5 were not measured, and less data were available for T6 because these vertebrae could not be clearly observed on lateral full spine X-ray. In addition, no patients with serious scoliosis (Cobb angle > 45°) were recruited because of the limited number of patients with severe AIS in the study. In future studies, patients with mild, moderate, and severe scoliosis should be recruited so that changes in vertebral body height can be observed at different severities of scoliosis, and the correlation between RASO and scoliosis severity can be further researched.