Influence of lumbar sagittal profile on pelvic orientation and pelvic motion during postural changes in patients with ankylosing spondylitis-related thoracolumbar kyphosis following pedicle subtraction osteotomy

OBJECTIVE The objective of this study was to investigate the impact of the lumbar sagittal profile on pelvic orientation and pelvic motion during postural changes in patients with ankylosing spondylitis (AS) and thoracolumbar kyphosis and to evaluate the potential risk of prosthetic dislocation after total hip arthroplasty (THA) following pedicle subtraction osteotomy (PSO). METHODS Seventy-two patients with AS-related thoracolumbar kyphosis following spinal osteotomy were retrospectively reviewed, and 21 healthy volunteers were recruited as a control group. Pre- and postoperative 2D full-body images in standing and sitting positions were obtained to evaluate the anterior pelvic plane angle (APPA), lumbar lordosis (LL), sacral slope (SS), pelvic tilt (PT), proximal femur angle (PFA), and femoroacetabular flexion during postural changes. Patients with AS were categorized in either a lordotic or kyphotic group based on the lumbar sagittal profile. RESULTS Significant increases in the SS and decreases in the APPA, PT, and LL were observed postoperatively in both the standing and sitting positions (p < 0.001 for all). Significantly higher APPA, PT, LL, and Delta PT, and lower SS, Delta SS, and Delta SS+Delta PFA were observed in the kyphotic group (p < 0.05). After undergoing PSO, Delta PT and Delta SS significantly decreased while femoroacetabular flexion significantly increased in both AS groups (p < 0.05), and no significant difference was present between the two groups (p > 0.05). Bath Ankylosing Spondylitis Radiology Hip Index scores in the kyphotic group were significantly worse than those in the lordotic group pre- and postoperatively (p < 0.05). No significant difference in parameters concerning pelvic motion (Delta APPA, Delta PT, and Delta SS) was found when PSO was performed in the thoracolumbar or lumbar spine. CONCLUSIONS Lumbar sagittal profiles greatly affect pelvic orientation and pelvic motion in AS. When THA is performed before PSO, AS patients with lumbar kyphosis are at higher risk of anterior prosthetic dislocation, while those with lordotic lumbar sagittal profiles are at higher risk of posterior dislocation. PSO should be performed prior to THA. After PSO, further decreased pelvic motion indicated a potential risk of posterior prosthetic dislocation after sequential THA, whereas theoretically patients with preoperative lumbar kyphosis are at higher risk of THA dislocation. The site where PSO was performed (thoracolumbar or lumbar spine) does not influence the risk of THA dislocation.