Introduction. Pedicle subtraction osteotomy is a powerful technique for correcting sagittal imbalance in ankylosing spondylitis. There has been significant perioperative morbidity associated with this technique in the peer review literature. We present the Royal National Orthopaedic Hospital experience with a single surgeon retrospective study that was conducted to evaluate the outcomes in patients who underwent lumbar pedicle subtraction osteotomy for the correction of thoracolumbar kyphotic deformity in ankylosing spondylitis. Method. Twenty seven patients underwent a lumbar pedicle subtraction osteotomy and adjacent level posterior instrumentation between 1995 and 2010. There were 18 males and 9 females in the study. Events during the peri-operative course and post-operative complications were recorded. The radiological outcome and patient satisfaction were analysed with mean follow-up of one and a half years. Results. The mean operative time was three and half hours and the mean blood loss was 2290mls. Final follow-up radiograph showed an increase in lumbar lordosis angle from 17 degrees to 45 degrees. The sagittal imbalance improved by 85mm with the operation. Complications included loosening in two patients, one transient neurologic deficit and one infective non-union occurred overall. There were no mortalities from the surgery. Two patients developed junctional kyphosis and required a repeat operation. There was an improvement in the Oswestry Disability Score from a mean of 29 to 16 after the surgey. All (100%) of the patients were satisfied with the results of the procedure and would recommend the surgey to others. Conclusion. The study shows that pedicle subtraction extension osteotomy is effective for the correction of kyphotic deformity in ankylosing spondylitis

Thoracic kyphosis increases with age. The resultant increase in compression forces on the anterior vertebral bodies leads to further kyphotic deformity and, an increased likelihood of vertebral collapse. This study aimed to determine the relative efficacy of two therapeutic strategies commonly used to treat hyperkyphosis. 69 subjects (26 male: 43 female) were randomised into 4 groups: strengthening, postural re-education (PEd), both and control. The strengthening group attended a gym 3 times a week for 12 weeks to perform seated extension exercises. The PEd group had 3 physiotherapy sessions within a 12 week period in which they received postural assessment and a home exercise programme. The combined group received both interventions while the control group received neither. Outcome measurements were assessed at baseline and 12 weeks. They included static (inclinometer) and 6-hour angular measurements (using flexible electrogoniometer (FEG)) and physical function tests. There were no significant differences between the marginal means of the angular measurements for any of the intervention groups. However, the group which received both interventions demonstrated reduced kyphosis as measured by the FEG angles (apex of the curve between T3 and T11), while the strengthening group showed reduced inclinometer angles (between T1 and T12). The strengthening group showed improvement in back extensor strength (BES) (0.6 +/− 0.2 N/kg, p < 0.01), time to walk 10 metres (−0.3 +/− 0.6 s, p < 0.05), and time to stand and sit 5 times (−0.9 +/− 0.6 s, p < 0.05). However, there was no relationship between change in BES and change in kyphotic angle. The PEd group showed the greatest improvement in the timed up and go test but this was not significant. Improvement in inclinometer angle over the 12 weeks was associated with degree of kyphosis at baseline (upright inclinometer r = −0.47, p=0.0001) but this relationship was not apparent in the FEG measurements. Both the FEG and inclinometer angles showed a marked decrease in degree of improvement in subjects aged >70. (50/50). A combination of strengthening and PEd was most effective at reducing hyperkyphosis. BES was improved with resisted strengthening but not with home-based postural exercises. However, increased BES was not associated with decreased kyphosis. Larger baseline kyphosis was associated with greater angular improvement. Subjects aged >70 were less likely to improve