Background Few studies on sagittal translation and its risk factors after pedicle subtraction osteotomy (PSO) in ankylosing spondylitis (AS) patients have been conducted. There is also no study on overall evaluation of radiologic parameters as the candidate of its risk factor.

Objectives The aim of this study was to report the cases of sagittal translation which developed after PSO in AS patients with kyphotic deformity and to analyze its risk factors

Methods The subjects of this study were 53 AS patients (58 cases) who underwent PSO to correct their kyphotic deformity between March 2006 and August 2016. The 53 subjects consisted of 45 males and 8 females. Their mean age was 39.3 12 Tit'>7.9 (range: 29–67). After osteotomy, the patient was examined for the presence of sagittal translation in the correction site through intraoperative radiograph. The low modified Stoke AS spine score (mSASSS) was measured before the surgery. The vertebral parameters such as lumbar lordosis angle, thoracic kyphotic angle, and sagittal vertical axis, and the pelvic parameters such as pelvic incidence, pelvic tilt, and sacral slope were also measured before and after the surgery.

The subjects were grouped according to the presence and absence of sagittal translation, and their radiologic parameters were compared. In addition, the correlation between sagittal translation and each parameter was analyzed. Complications that developed during and after the surgery were also analyzed.

Results Sagittal translation developed in 16 subjects (30%) or 17 cases (29.3%). The mean lumbar lordosis angle and the mean sagittal vertical axis of both the sagittal translation (ST) group and the non-sagittal translation (Non-ST) group were successfully corrected (p=0.000, respectively). A significant difference in preoperative mean sacral slope was observed between the groups (p=0.045). The ST group showed a significantly higher mSASSS (48.1±20.7) than the Non-ST group (36.8±16.2) (p=0.002). In the multivariate regression analysis, sagittal translation was positively correlated with mSASSS (odds ratio 1.34, P=0.002) and the preoperative sacral slope (odds ratio 1.46, P=0.009), and negatively correlated with the difference between preoperative and postoperative thoracic kyphotic angle (odds ratio 0.68, P=0.01). Both groups showed no finding of permanent neurologic complication after the surgery.

Conclusions The incidence of sagittal translation after pedicle subtraction osteotomy was closely related with the severity of ankyloses in AS patients. Therefore, when pedicle subtraction osteotomy is performed for AS patients with severe ankyloses and high sacral slope, it is required that surgeon consider sagittal translation which could induce neurologic complication.

Disclosure of Interest None declared