Computed tomography-guided navigation of thoracic pedicle screws for adolescent idiopathic scoliosis results in more accurate placement and less screw removal

Study Design.: Retrospective study of computed tomography-guided navigation (CTGN) of thoracic pedicle screw placement in patients with adolescent idiopathic scoliosis (AIS). Objective.: To compare the accuracy and safety of thoracic pedicle screw placement and frequency of intraoperative removal using CTGN versus conventional freehand technique in AIS. Summary of Background Data.: Even in experienced hands, more than 10% of the thoracic pedicle screws are misplaced. CTGN may improve accuracy and safety, but there is little published data on its efficacy. Methods.: We reviewed intraoperative computed tomographic images in a consecutive series of AIS cases undergoing posterior fusion during a 1-year period. Three types of screws were identified: an optimal screw-the central axis is in the plane and axis of the pedicle with the tip completely within the vertebral body; an acceptable screw-the majority of its shank is outside the central axis of the pedicle, but not potentially unsafe; and a potentially unsafe screw-(1) the central axis of the screw traversed the canal, (2) left anterior/lateral vertebral body perforation, risking the aorta, or (3) any screw repositioned or removed after the postimplant computed tomography. Results.: In 42 patients, 485 screws were evaluable with a visible pedicle and screw (300 navigated and 185 non-navigated). Screws were classified as follows: optimal screws, 74% CTGN versus 42% non-navigated; acceptable screws, 23% CTGN versus 49% non-navigated; and potentially unsafe, 3% CTGN versus 9% non-navigated (P < 0.001). A potentially unsafe screw was 3.8 times less likely to be inserted with navigation (P=0.003). The odds of a significant medial breach were 7.6 times higher without navigation (P < 0.001). A screw was 8.3 times more likely to be removed intraoperatively in the non-navigated cohort (P=0.003). Conclusion.: CTGN resulted in more optimally placed thoracic pedicle screws, fewer potentially unsafe screws, and fewer screw removals.