Reliability and consequences of intraoperative 3D imaging to control positions of thoracic pedicle screws

Abstract

The insertion of thoracic pedicle screws (T1-T10) is subject to a relevant rate of malplacement. The optimum implantation procedure is still a topic of controversial debate. Currently, a postoperative computed tomography is required to evaluate the screw positions. The present study was undertaken to clarify whether intraoperative 3D imaging is a reliable method of determining the position of thoracic pedicle screws. This prospective study involved 40 consecutive patients with thoracic spinal injuries, with intraoperative 3D scans being performed to determine the positions of 240 pedicle screws in T1-T10. The results of the 3D scans were compared with the findings of postoperative CT scans, using a clinical classification system. The positions of 204 pedicle screws could be viewed by means of both 3D and CT scans and the results compared. The 3D scans achieved a sensitivity of 90.9 % and a specificity of 98.8 %. The rate of misclassification by the 3D scans was 2.5 %. Nine pedicle screws were classified as misplaced and their position corrected intraoperatively (3.8 %). No screws required postoperative revision. Performing an intraoperative 3D scan enables the position of thoracic pedicle screws to be determined with sufficient accuracy. The rate of revision surgery was reduced to 0 %.