The effects of scoliosis and subsequent surgery on the shape of the torso

Abstract

BackgroundAdolescent idiopathic scoliosis (AIS) causes asymmetry of the torso, and this is often the primary concern of patients. Surgery aims to minimise the visual asymmetry. It is not clear how scoliosis makes the torso asymmetric or how scoliosis surgery changes that asymmetry when compared to the distribution of asymmetries seen in a non-scoliotic group of normal controls.MethodsSurface topography images were captured for a group with AIS both pre-operatively and post-operatively. Identifiable points were compared between the images to identify the effects of AIS on the shape of the torso by looking at the relative heights and distances from the midline of the shoulders, axillae and waist in a two-dimensional coronal view. This was then compared to a previously reported group of normal non-scoliotic children to analyse whether surgery recreated normality.ResultsThere were 172 pairs of images with 164 females and 8 males, mean age at pre-operative scan of 13.7 years. The normal group was 642 images (237 females and 405 males) from 116 males and 79 females, mean age of 12.5 years.The curve patterns seen in the scoliotic group matched the patterns of a main thoracic curve (n = 146) and main thoracolumbar curve (n = 26). The asymmetries seen in both shoulders, axillae and waist were different between the two different types of curve. Across both groups, the shoulder asymmetry was less than that of the corresponding axillae.There was a statistically significant reduction in all asymmetries following surgery in the main thoracic group (p < 0.001). This was not seen in the main thoracolumbar group, thought to be due to the small sample size. In the main thoracic group, there were statistically significant differences in the asymmetries between the post-operative and normal groups in the shoulders and axillae (p < 0.001) but not the waist.ConclusionsThis paper demonstrates quantitatively the range of asymmetries seen in the AIS torso and the degree to which surgery alters them. Surgery does not recreate normality but does cause a statistically significant change in torso shape towards that seen in a non-scoliotic group.

A Description of Three-Dimensional Shape of the Posterior Torso Comparing Those with and without Scoliosis

“…The origin of the axes is at the vertebra prominens in all raw data sets. As had been previously discovered from these data sets, there were two obvious patterns to the data [4,5]. Thus, prior to analysis, both the non-scoliotic and scoliotic cohorts were subdivided into a main thoracic curve pattern and a main thoracolumbar curve pattern (defined by the location of the largest curve) [8].…”

“…Patients and surgeons are concerned about the spine and the asymmetry of the torso [2,3,18]. Previous literature has identified the variability of the posterior torso shape in both a non-scoliotic cohort [4] and scoliotic cohort [5]. These papers examined the pre-operative and post-operative shape in comparison to the non-scoliotic cohort using data ellipses.…”

“…For both non-scoliotic and scoliotic cohorts, the subjects were sub-divided to allow for different curve types [8]. This again followed previous methodology [5] and was to ensure that any variability of shape seen between different curve types was identified. In the non-scoliotic cohort, there was a statistically significant but clinically insignificant difference in the mean shapes between males and females [16].…”

“…For the scoliotic cohort, the small numbers of males when compared to the number of females made the analysis by sex unreliable. Thus, following the methodology that has already been established in the literature by Gardner et al [5], further analysis was performed independent of sex.…”

“…This visible asymmetry is of concern to patients [1] and is assessed by physicians as part of the physical examination [2,3]. Previous work using data ellipses [4,5] has analysed two-dimensional (2D) torso shape and defined the mean values and variability of the position of different torso anatomical points in both a non-scoliotic and a pre-operative and postoperative scoliotic cohort. However, the overall 3D shape of the torso in the non-scoliotic cohort and the scoliotic cohort, both pre-operatively and post-operatively, has not been previously assessed.…”