Reliability of three-dimensional fluoroscopy for detecting pedicle screw violations in the thoracic and lumbar spine

OBJECTIVE: Thoracic and lumbar pedicle screws have become popular because of their biomechanical superiority over other methods of spinal fixation. However, the safety and efficacy of transpedicular screws depend on their proper placement. Recent advances in imaging have resulted in the ability to acquire three-dimensional (3-D) axial images of the spine during surgery, and this study was undertaken to assess the reliability of this technology to detect,pedicle violations. METHODS: Pedicle screws were placed in six human cadaver spines from T1 to S1 using standard techniques. Intentional pedicle violations were created in 74 of 216 pedicles, and violations were graded on a four-point scale (range, 0-3). Radiographic images were then obtained using a conventional spiral computed tomographic scanner and the Siremobil Iso-C 3D (Siemens Medical Solutions, Erlangen, Germany) 3-D fluoroscopy unit. An independent neuroradiologist then graded pedicle violations as ascertained by the two imaging modalities. RESULTS: Using direct inspection of the pedicles as the "gold standard," the overall sensitivity and specificity for detecting pedicle violations were 0.716 and 0.789, respectively, with 3-D fluoroscopy. The overall sensitivity and specificity for detecting pedicle violations were 0.608 and 0.937, respectively, with conventional computed tomography. All Grade 2 pedicle violations were detected in the thoracic spine by both modalities, and all Grade-3 violations were detected by both modalities. CONCLUSION: Axial images obtained with 3-D fluoroscopy demonstrate a higher sensitivity but lower specificity than conventional computed tomographic scanning for assessing pedicle violations. By providing real-time intraoperative imaging, 3-D fluoroscopy may enhance the safety of thoracic transpedicular instrumentation.