Does the preoperative lumbar sagittal profile affect the selection of osteotomy level in pedicle subtraction osteotomy for thoracolumbar kyphosis secondary to ankylosing spondylitis?

Objective: To investigate the different preoperative lumbar sagittal profiles of ankylosing spondylitis (AS) patients and the selection of osteotomy level for one-level pedicle subtraction osteotomy (PSO) for the correction of thoracolumbar kyphosis.

Patients and methods: Seventy-one consecutive AS patients with an average age of 35.3 years and a mean follow-up time of 35.9 months who underwent one-level PSO for thoracolumbar kyphosis were divided into 2 groups based on their preoperative lumbar sagittal profiles as follows: group A, lordotic lumbar sagittal profiles; and group B, kyphotic lumbar sagittal profiles. The following radiological parameters were measured and compared: chin-brow vertical angle (CBVA), global kyphosis (GK), thoracic kyphosis (TK), lumbar lordosis (LL), sagittal vertical axis (SVA), pelvic incidence (PI), pelvic tilt (PT) and sacral slope (SS). Clinical evaluation included Oswestry Disability Index (ODI) and Visual Analogue Scale (VAS). Perioperative and mid-term complications were reviewed.

Results: There were 28 patients in group A and 43 in group B. The preoperative LL was -21.0° in group A and 2.3° in group B (P < 0.05). The preoperative SVA was 122.5 mm in group A and 184.3 mm in group B (P < 0.05). All the patients in group A (100%) underwent PSO at L1/L2, while 90% of group B patients underwent PSO at L2/L3, with no significant difference of postoperative GK, LL and SVA between the 2 groups (P > 0.05). No obvious loss of correction was observed in either group at the final follow-up. The correction of LL and SVA showed a strong but not statistically significant increasing trend as the PSO level descended from L1 to L3 (P > 0.05). The postoperative ODI was significantly lower in patients underwent PSO at L1 or L2 (P < 0.05).

Conclusions: Patients in group B had significantly worse preoperative sagittal alignments compared to group A. The distribution of osteotomy levels varied between the 2 groups due to the different lumbar profiles; however, satisfactory correction was achieved in both groups. The preoperative lumbar profiles need to be considered in selecting the optimal osteotomy level. Patients with kyphotic lumbar profiles are suitable candidates for PSO at L2/L3, while L1/L2 PSO is appropriate for patients with lordotic lumbar profiles.