Intraoperative computed tomography image-guided navigation for posterior thoracolumbar spinal instrumentation in spinal deformity surgery

Abstract

Object: Placement of thoracolumbar pedicle screws in spinal deformity surgery has a reported inaccuracy rate as high as 30%. At present, image-guided navigation systems designed to improve instrumentation accuracy typically use intraoperative fluoroscopy or preoperative CT scans. The authors report the prospective evaluation of the accuracy of posterior thoracolumbar spinal instrumentation using a new intraoperative CT operative suite with an integrated image guidance system. They compare the accuracy of thoracolumbar pedicle screw placement using intraoperative CT image guidance with instrumentation placement utilizing fluoroscopy.

Methods: Between December 2007 and July 2008, 12 patients underwent posterior spinal instrumentation for spinal deformity correction using intraoperative CT-based image guidance. An intraoperative CT scan of the sterile surgical field was obtained after decompression and before instrumentation. Instrumentation was placed, and a postinstrumentation CT scan was obtained before wound closure to assess the accuracy of instrumentation placement and the potential need for revision. The accuracy of pedicle screw placement was later reviewed and recorded by independent observers. A comparison group of 14 patients who underwent thoracolumbar instrumentation utilizing fluoroscopy and postoperative CT scanning during the same time period was evaluated and included in this analysis.

Results: In the intraoperative CT-based image guidance group, a total of 164 thoracolumbar pedicle screws were placed. Two screws were found to have breached the pedicle wall (1.2%). Neither screw was deemed to need revision due to misplacement. In the comparison group, 211 pedicle screws were placed. Postoperative CT scanning revealed that 11 screws (5.2%) had breached the pedicle. One patient in the fluoroscopy group awoke with a radiculopathy attributed to a misplaced screw, which required revision. The difference in accuracy was statistically significant (p = 0.031).

Conclusions: Intraoperative CT-based image guidance for placement of thoracolumbar instrumentation has an accuracy that exceeds reported rates with other image guidance systems, such as virtual fluoroscopy and 3D isocentric C-arm-based stereotactic systems. Furthermore, with the use of intraoperative CT scanning, a postinstrumentation CT scan allows the surgeon to evaluate the accuracy of instrumentation before wound closure and revise as appropriate.