The influence of thoracic inlet angle on the cervical sagittal alignment in ankylosing spondylitis patients with thoracolumbar kyphosis following lumbar pedicle subtraction osteotomy

English Keywords:Ankylosing spondylitis  Thoracolumbar kyphosis  Pedicle subtraction osteotomy  Thoracic sagittal alignment  Cervical sagittal alignment

English Abstract:

  【Abstract】 Objectives: To investigate the radiographic characteristics of cervical sagittal alignment in ankylosing spondylitis (AS) patients with thoracolumbar kyphosis at different thoracic inlet angles(TIA). Methods: 32 patients who underwent pedicle subtraction osteotomy (PSO) in our hospital from January 2012 to December 2020 with complete data were included. Clinical data including gender, age, osteotomized level, fusion levels were collected. Preoperative, postoperative, the last follow-up radiological parameters consisting of cervical sagittal vertical axis(cSVA), McGregor slope(McGs), upper cervical lordosis(UCL), lower cervical lordosis(LCL), cervical lordosis(CL), C2 slope, C7 slope, T1 slope(T1S), neck tilt(NT), thoracic inlet angle(TIA), global kyphosis(GK), thoracic kyphosis(TK), lumbar lordosis(LL), pelvic incidence(PI), pelvic tilt(PT), sacral slope(SS), and local kyphosis(LK) were measured. Patients were divided into two groups on the basis of TIA value (Group A: TIA<70°; Group B: TIA≥70°). Independent sample t-test or chi-square test was used to compare the difference of clinical data and preoperative parameters between the two groups, and paired sample t-test was used for the comparison of preoperative and postoperative parameters in group A and group B respectively. Pearson correlation analysis was performed to evaluate the correlations between TIA and clinical data and preoperative parameters. Results: There were 15 patients (13 males and 2 females) in group A and 17 patients (13 males and 4 females) in group B. The mean age in group A was 28.5±5.7 years, and that in group B was 37.7±14.4 years. There was significant difference in the age between the two groups(P<0.05). The preoperative cSVA, UCL, CL, TK and GK in group B were 48.0±18.1mm, -22.8°±7.4°, -56.3°±10.9°, 50.1°±9.9° and 69.8°±11.9° respectively, which were greater than those in group A(P<0.05). While no significant difference was observed in preoperative McGs and LCL between the two groups(P>0.05). The postoperative LCL, CL, TK in group B were -20.9°±11.9°, -35.0°±9.4° and 46.2°±12.0° respectively, which were greater than in group A. While there was no significant difference being observed in postoperative cSVA, McGS, UCL and NT between the two groups(P>0.05). The loss of correction at the last follow-up showed no significant difference between the two groups(P>0.05). Preoperative TIA was positively correlated with age(r=0.549, P=0.001), cSVA(r=0.367, P=0.039), TK(r=0.392, P=0.027) and GK(r=0.366, P=0.039). However, TIA was negatively correlated with preoperative UCL(r=-0.591, P<0.001) and CL(r=-0.357, P=0.045). Moreover, no correlation between the change of cervical lordosis and ΔLK was found in both groups(P>0.05). Conclusions: Compared with AS patients with small TIA, AS patients with greater TIA had larger preoperative UCL, whereas there was no difference in LCL. After lumbar PSO, both UCL and LCL decreased. However, AS patients with greater TIA had larger LCL and there was no significant difference concerning UCL.