Sergei Vissarionov1, Joshua E Schroeder2, Sergey N Novikov3, Dmitriy Kokyshin4, Sergei Belanchikov4, Leon Kaplan5. 1. Department of Science, Federal State Budgetary Institution Turner Research Institute of Infantile Orthopedics, Russian Ministry of Health and Social Development, St. Petersburg, Russia 196603; Department of Pediatric Traumatology and Orthopedics, SEI FVE Saint-Petersburg Medical Academy of Postgraduate Education, 64-68 Parkovaya Street, Pushkin, St. Petersburg, Russia 196603. 2. Department of Spine Surgery, Hospital for Special Surgery, 535 East 70th st, New York, NY 10021 USA. 3. Department Radiation Oncology and Nuclear Medicine, N.N. Petrov Institute of Oncology, 68 Leningradskaya St, St Petersburg, Russia. 4. Department of Spinal Disorders and Neurosurgery, Federal State Budgetary Institution Turner Research Institute of Infantile Orthopedics, Russian Ministry of Health and Social Development, 64-68 Parkovaya Street, Pushkin, St. Petersburg, Russia 196603. 5. Spine Surgery Unit, Department of Orthopedic Surgery, Hadassah-Hebrew University Medical Center, POB 12000, Jerusalem 91120, Israel. Electronic address: kapileon@netvision.net.il.
Abstract
STUDY DESIGN: Ambispective study of patients undergoing surgical correction of adolescent idiopathic scoliosis. OBJECTIVE: To evaluate the accuracy of screw placement using preoperative 3-dimensional (3D) computed tomography (CT)-based navigation with intraoperative fluoroscopic guidance compared with freehand placement. SUMMARY OF BACKGROUND DATA: Pedicle screws placed in deformed vertebrae have a high malposition rate. The use of navigation-based systems has increased placement accuracy. METHODS: Intraoperative registration of patient anatomy to preoperative 3D-CT was performed using anatomic landmarks. When registration accuracy was high (mean square error, <1.0 mm), screw tracts were drilled under navigation guidance; when the error was >1.0 mm, re-registration was performed. The researchers documented times for registration, navigation, and screw placement, and the number of passes. Results were compared with outcomes in cases operated on with freehand screw placement. RESULTS: A total of 62 patients were included (54 females and 8 males; mean age was 15.1 years [range, 12-18 years]). Mean deformity was 67° (range, 52° to 80°). Mean follow-up was 35 months (range, 42-19 months). In the navigation group, 710 pedicle screws were placed. Mean times were 55 seconds for tracker placement, 94.5 seconds per vertebra for patient registration, 131.1 seconds for screw tract formation on the concave side of the deformity, and 129.5 seconds on the convex side. Average total procedure time was 3.5 hours (range, 2-7 hours). Mean overall registration accuracy was 0.7 mm. Pedicle integrity was breached in 1.6% trajectories. In the freehand group, 470 pedicle screws were placed. Average time for screw placement was 135.2 seconds (p < .001 vs. navigation). Pedicle integrity was breached in 5.1% of trajectories (p < .0001 vs. navigation). No patient developed neurological or other complications. There was no destabilization of the instrumented spine during short- or long-term follow-up. CONCLUSIONS: Intraoperative optic fluoroscopic navigation based on anatomic landmark registration to preoperative 3D-CT spine images enables precise pedicle screw placement with only a minor increase in pedicle preparation time in patients with adolescent idiopathic scoliosis.
STUDY DESIGN: Ambispective study of patients undergoing surgical correction of adolescent idiopathic scoliosis. OBJECTIVE: To evaluate the accuracy of screw placement using preoperative 3-dimensional (3D) computed tomography (CT)-based navigation with intraoperative fluoroscopic guidance compared with freehand placement. SUMMARY OF BACKGROUND DATA: Pedicle screws placed in deformed vertebrae have a high malposition rate. The use of navigation-based systems has increased placement accuracy. METHODS: Intraoperative registration of patient anatomy to preoperative 3D-CT was performed using anatomic landmarks. When registration accuracy was high (mean square error, <1.0 mm), screw tracts were drilled under navigation guidance; when the error was >1.0 mm, re-registration was performed. The researchers documented times for registration, navigation, and screw placement, and the number of passes. Results were compared with outcomes in cases operated on with freehand screw placement. RESULTS: A total of 62 patients were included (54 females and 8 males; mean age was 15.1 years [range, 12-18 years]). Mean deformity was 67° (range, 52° to 80°). Mean follow-up was 35 months (range, 42-19 months). In the navigation group, 710 pedicle screws were placed. Mean times were 55 seconds for tracker placement, 94.5 seconds per vertebra for patient registration, 131.1 seconds for screw tract formation on the concave side of the deformity, and 129.5 seconds on the convex side. Average total procedure time was 3.5 hours (range, 2-7 hours). Mean overall registration accuracy was 0.7 mm. Pedicle integrity was breached in 1.6% trajectories. In the freehand group, 470 pedicle screws were placed. Average time for screw placement was 135.2 seconds (p < .001 vs. navigation). Pedicle integrity was breached in 5.1% of trajectories (p < .0001 vs. navigation). No patient developed neurological or other complications. There was no destabilization of the instrumented spine during short- or long-term follow-up. CONCLUSIONS: Intraoperative optic fluoroscopic navigation based on anatomic landmark registration to preoperative 3D-CT spine images enables precise pedicle screw placement with only a minor increase in pedicle preparation time in patients with adolescent idiopathic scoliosis.