OBJECTIVETo investigate the feasibility and safety of the treatment for thoracolumbar fractures with pedicle screw at the fracture level and vertebroplasty via paraspinal approach.

METHODSFrom August 2007 to August 2010, 22 old patients with thoracolumbar fractures were treated with pedicle screw at the fracture level and vertebroplasty via paraspinal approach. There were 14 males and 8 females, ranging in age from 60 to 71 years (mean, 64.6 years). The time from injury to surgery varied from 1 to 4 d (mean,2.7 d). All the patients suffered from single thoracolumbar fractures and located at T11 in 2 cases, at T12 in 5 cases, at L1 in 11 cases and at L2 in 4 cases. According to the Denis fracture classification, there were 6 compression fractures and 16 burst fractures. The mean preoperative load-sharing classification of spine fractures was 5.4 score. The mean preoperative thoracolumbar injury classification and scoring was 5.2. Based on the ASIA neurologic grading system, preoperative neurological function was grade B in 2 cases,grade C in 3 cases, grade D in 7 cases and grade E in 10 cases. The neurological function, vertebral central and anterior height, kyphotic angle of the vertebral fractures by radiographs and visual analog scale were calculated pre-operatively, post-operatively and at the last follow-up.

RESULTSMedian operating time was 60.8 min (ranged from 50 to 95 min) and median blood loss was 84 ml (ranged from 50 to 130 ml). The operative incisions were healed well. The duration of follow-up averaged 21.6 months (ranged from 12 to 48 months). The anterior vertebral body height was corrected from preoperative (52.3 +/- 10.3) % to postoperative (6.1 +/- 4.2) % and (6.8 +/- 5.4) % at the last follow-up. The central vertebral body height was corrected from preoperative (38.9 +/- 11.2) % to postoperative (8.3 +/- 4.7) % and (9.4 +/- 4.5)% at the last follow-up. The Cobbs angle of the injured vertebral segment was corrected from preoperative (19.5 +/- 9.5) degrees to postoperative (4.3 +/- 4.1) degrees and (6.2 +/- 4.7) degrees at the last follow-up. The VAS scores reduced from preoperative 8.56 +/- 0.88 to post-operative 3.48 +/- 0.91 and 3.20 +/- 0.92 at the last follow-up. The postoperative neurologic function of all 22 patients improved 1 to 2 degrees except 10 patients of grade E. There were no instances of instrumentation failure and no patient had persistent postoperative back pain.

CONCLUSIONThe pedicle screw at the fracture level and vertebroplasty via paraspinal approach has the advantages of less invasive and blood loss, and could prevent the development of kyphosis and offers improvement of the spinal cord function. Furthermore, it could decrease the risks of postoperative back pain and the failure of instrumentation.