Objective: Pedicle screw fixation in the lumbar spine has become the standard of care for various causes of spinal instability. However, because of the smaller size and more complex morphology of the thoracic pedicle, screw placement in the thoracic spine can be extremely challenging. In several published series, cortical violations have been reported in up to 50% of screws placed with standard fluoroscopic techniques. The goal of this study is to evaluate the accuracy of thoracic pedicle screw placement by use of image-guided techniques.
Methods: During the past 4 years, 266 image-guided thoracic pedicle screws were placed in 65 patients at the University of Michigan Medical Center. Postoperative thin-cut computed tomographic scans were obtained in 52 of these patients who were available to enroll in the study. An impartial neuroradiologist evaluated 224 screws by use of a standardized grading scheme. All levels of the thoracic spine were included in the study.
Results: Chart review revealed no incidence of neurological, cardiovascular, or pulmonary injury. Of the 224 screws reviewed, there were 19 cortical violations (8.5%). Eleven (4.9%) were Grade II (< or =2 mm), and eight (3.6%) were Grade III (>2 mm) violations. Only five screws (2.2%), however, were thought to exhibit unintentional, structurally significant violations. Statistical analysis revealed a significantly higher rate of cortical perforation in the midthoracic spine (T4-T8, 16.7%; T1-T4, 8.8%; and T9-T12, 5.6%).
Conclusion: The low rate of cortical perforations (8.5%) and structurally significant violations (2.2%) in this retrospective series compares favorably with previously published results that used anatomic landmarks and intraoperative fluoroscopy. This study provides further evidence that stereotactic placement of pedicle screws can be performed safely and effectively at all levels of the thoracic spine.