Complications of Thoracic Pedicle Screws in Scoliosis Treatment

Abstract
Study Design. A retrospective study. Objective. To analyze complications with thoracic pedicle screws in scoliosis treatment at our Department over a 3-year period (1999–2001). Summary of Background Data. The use of pedicle screws remains controversial for thoracic scoliosis for fear of complications. Methods. A total of 115 consecutive patients who underwent posterior fusion using 1035 transpedicular thoracic screws were reviewed. All patients presented a main thoracic scoliosis with a mean Cobb angle of 75.4° (range, 60°–105°). For thoracic screw placement, a mini-laminotomy technique was used, inserting a spatula inside the vertebral canal to palpate the borders of the pedicle. Postoperative CT scan was used in 25 patients (21.7%) to study a total of 311 screws, when the screw position was questionable. Results. An independent spine surgeon retrospectively reviewed medical records and radiographs of the patients, at a mean follow-up of 4 years. There were 18 screws misplaced (1.7%) in a total of 13 patients (11.3%). Screw malposition was symptomatic only in 1 patient (pleural effusion and fever) and asymptomatic in the other 12 cases (10.4%). Other complications included intraoperative pedicle fractures in 15 patients (13%), dural tears (without neurologic complications) in 14 cases (12.1%), and superficial wound infections in 2 (1.7%). Another operation for screw removal was performed in 5 patients (4.3%), due to pleural effusion (in 1 case), asymptomatic late lateral loosening of a malpositioned screw (in 1), and the possible future risks related the intrathoracic screw position despite the lack of any symptoms (in 3). Two cases (1.7%) were retreated due to wound infection, without removing instrumentation. There was no loss of correction at follow-up. Conclusions. The thoracic pedicle screw placement in scoliosis patients requires utmost caution. The mini-laminotomy technique was beneficial in increasing safety of the procedure with an acceptable incidence of complications.