Posterior vertebral column resection in fixed lumbosacral deformity

Study Design. Retrospective study. Objectives. To report the results and the technique of posterior vertebral column resection (PVCR) in fixed lumbosacral deformity. Summary of Background Data. Fixed lumbosacral deformity has gross imbalance as well as progressive compensatory thoracolumbar deformity because of the absence of a mobile spine caudally. There is no consensus of the surgical treatment of fixed lumbosacral deformity. Fixed lumbosacral deformity can be managed by PVCR, which not only corrects the deformity but also restores the trunk balance by translation of spinal column. Methods. Twenty-five consecutive fixed lumbosacral deformity patients subjected to PVCR were reviewed after a minimum follow-up of 2 years. The apical vertebra and the ultimately resected vertebra were below L4 in most cases. Etiologic diagnoses were congenital scoliosis in 6, congenital kyphoscoliosis in 3, post-traumatic kyphosis in 2, and postinfectious kyphosis in 14. The average age at the operation was 38 years (range, 2.5 - 61 years) with a male: female ratio of 7: 18. The indication for PVCR was fixed lumbosacral deformities that could not be brought to a reasonable balance on traction or forced side bending. Before surgery, 12 patients (48%) presented with neurologic compromise and 20 patients (80%) presented with intractable back pain. The surgical technique consisted of segmental pedicle screw fixation, decompression of neural structures, and resection of the vertebral column at the apex of the deformity via the posterior route, followed by deformity correction and global fusion. During the resection, all the nerve roots were preserved. Results. Two and one tenth (2.1) vertebrae removed on average (range, 1 - 5 vertebrae), and 52 in all. The average fusion extent was 4.5 vertebrae (range, 2 - 8 vertebrae). The anterior column reconstruction was carried out with autogenous bone graft in all patients and additional insertion of titanium mesh in 12. The distal anchor went down to L5 in 4 patients, S1 in 12, and S2 in 9. Preoperative scoliosis of 38 degrees +/- 12 degrees was corrected to 15 degrees +/- 8 degrees (60% correction) at most recent follow-up, and preoperative kyphosis of 35 degrees +/- 25 degrees was corrected to -5 degrees +/- 11 degrees (40 degrees correction). The compensatory coronal curve of 26 degrees was spontaneously corrected to 12 degrees at most recent follow-up. The preoperative thoracic lordosis of -16 degrees was corrected to 4 degrees at most recent follow-up, and the preoperative lumbar kyphosis of 20 degrees was corrected to -17 degrees. Preoperative coronal imbalance of 2.0 cm was improved to 0.9 cm at most recent follow-up, and preoperative sagittal imbalance of 9.3 cm was improved to 4.6 cm. Mean operation time was 280 minutes with a blood loss of 2,810 mL. The preoperative neurologic compromise was improved in all patients, and pain was reduced to less than half of preoperative visual analog analog scale in 17 patients. The following complications were encountered in 5 patients: 2 with transient neurology that spontaneously improved without a sequel within 6 months, 2 with compression fractures at proximal adjacent vertebra, and 1 with pseudarthrosis. Conclusions. PVCR is an effective procedure for the management of fixed lumbosacral deformity. It provides satisfactory correction and improved functional outcomes. However, it is a technically demanding and exhausting procedure with possible risks for complications.