Comparative Analysis of Global Spine Alignment using Whole Spine CT and X-Ray: Correlative Analysis of Reciprocal and Physiologic Changes of the Cervical Alignment

Abstract

Objective The aims of this study were to compare the parameters of spinal sagittal alignment between whole spine computed tomography (CT) scans in supine position and whole spine X-ray in standing position, and also to analyze the reciprocal and physiologic changes of the cervical spine in dynamic motion. Methods Data were obtained retrospectively in a single hospital. Radiographic data of 55 subjects were recruited about whole spine CT, whole spine X-ray, and cervical dynamic X-ray from November 2007 to May 2014. The mean age was 52.8 year (range, 17–85 years) and the female:male ratio was 31:24. Whole spine CT was checked in supine position and whole spine X-ray was checked in standing position. Cervical dynamic X-ray was checked in standing position. We included the subjects only with mild degenerative disease and simple back pain and excluded the subjects with severe scoliosis, severe deformity, and operation history. The radiological sagittal parameters on lateral paragraph were evaluated including C0–C7 Cobb angle (C0C7), C0–C2 Cobb angle (C0C2), C2–C7 Cobb angle (C2C7), T1 slope (T1S), neck tilt (NT), thoracic inlet angle (TIA), C2 tilt (C2T), C7 tilt (C7T), C2 sagittal vertical axis (C2SVA), C7 sagittal vertical axis (C7SVA), T1–T12 Cobb angle (T1T12), T1–T7 Cobb angle (T1T7), T7–T12 Cobb angle (T7T12), L1–S1 Cobb angle (L1S1), sacral slope (SS), pelvic tilt (PT), and pelvic incidence (PI). Each parameter was analyzed statistically with paired t test between CT and X-ray, and also assessed the relationships between all parameters with the Pearson correlation analysis. Results All corresponding parameters of spine alignment on X-ray were significantly correlated with those on CT. There was no significant difference in TIA (0.62 ± 3.41) and PI (0.62 ± 3.55) between CT and X-ray ( p > 0.05). However, cervical lordosis, T1 slope (T1S), thoracic kyphosis, and lumbar lordosis (L1S1) on X-ray were significantly larger than those on CT. The differences of the means were 21.3 ± 15.6 degrees in C0C7, 3.8 ± 9.5 degrees in C0C2, 17.5 ± 12.3 degrees in C2C7, 7.8 ± 11.3 degrees in T1S, 14.3 ± 9.8 degrees in T1T12, 9.1 ± 8.4 degrees in T1T7, 5.2 ± 9.0 degrees in T7T12, and 4.8 ± 17.1 degrees in L1S1 ( p < 0.05). Age significantly correlated with TIA ( r = 0.382). Thoracic kyphosis (T1T12) had the significant correlation with T1S ( r = 0.874 on CT, r = 0.425 on X-ray) and cervical lordosis (C0C7; r = 0.455 on CT, r = 0.405 on X-ray). Cervical tilt (C2T) and SVA (C2SVA) also had the significant correlations with T1S (C2T: r = − 0.570 on CT, r = − 0.550 on X-ray; C2SVA: r = 537 on CT, r = 0.498 on X-ray). T1S had the significant correlations with the TIA ( r = 0.414 on CT, r = 0.368 on X-ray). T1S also had the significant correlations with cervical lordosis (C0C7: r = 0.367 on CT, r = − 0.469 on X-ray), which mainly correlated with subaxial spine (C2C7: r = 0.355 on CT, r = 0.404 on X-ray) and not with upper cervical spine (C0C2). The changes of the cervical lordosis (C0C7) in dynamic motion (neutral, flexion, and extension) had highly evident correlations with subaxial spine (C2C7) than with upper cervical spine (C0C2). Pearson correlation coefficients ( R) between C0C7 and C2C7 were 0.810 in neural, 0.766 in flexion, and 0.740 in extension, respectively ( p < 0.01). Conclusion The difference of spine alignment between CT and X-ray should be because of the difference of the gravity stem from the posture. Upright posture may increase thoracic kyphosis, and cervical lordosis sequentially and reciprocally. Physiologic motions in the spine cause the cervical accommodations, in which subaxial spine is mainly and initially responsible for the reciprocal changes to other spinal parameters such as TIA and T1 slope. This finding may be the main reason of the subaxial cervical spine degeneration.