[3D-Navigated Implantation of Pedicle Screws in the Cervical Spine - Experience and Analysis of Complications]

Abstract

Background: Placing transpedicular screws in the cervical spine is a special challenge for spine surgeons, due to the anatomical features of this part of the spine. During the last 15 years, computer-aided navigation systems have been developed to facilitate this procedure and to make it safer for patients. One option is navigation by intraoperatively acquired data sets with the use of an 3D C-arm. Patients/Material and methods: Our retrospective study evaluates transpedicular screws in the cervical spine placed by 3D C-arm navigation, within a 6 year period in a level 1 trauma centre. We recorded epidemiological data, operation time and especially general adverse events, as well as revision surgery, including reasons for revision. We used a C-arm Arcardis Orbic 3D (Siemens, Munich), connected to a navigation system (VectorVision, Brainlab, Munich). Results: Between July 2007 and July 2013, 207 transpedicular screws were placed in 58 patients. The main indications were trauma (69 %), rheumatic diseases (20.7 %) and tumour (8.6 %). The most commonly instrumented cervical spine segments were C2 (53.5 %)%), C7 (10.3 %) and C5 (8.6 %). In nearly 95 % of the cases, we performed an intraoperative 3D scan after screw or k-wire placement to control the screw position. We found unacceptable malposition in 7.2 % of patients. This was corrected at once. Ten patients had to be revised: seven times due to wound problems, twice because of implant failure and once for treatment of CSF leakage. Three screws (1.5 %) led to injuries of the vertebral artery, once with a lethal outcome. Analysis of these cases showed that the 3D scan gave reduced data quality, due to reduced bone density or anatomical factors. Conclusion: Intraoperative 3D C-arm navigation seems to be a reliable option for transpedicular screw placement in the cervical spine. Complication rates were comparable to published values. 7.2 % of all screws were corrected intraoperatively after a control scan. Therefore possible revisions could be avoided during primary surgery. Analysis of problematic cases led to a change in our treatment strategy: in patients with poor bone quality and/or anatomical problems which lead to 3D scans of poor quality, we avoid transpedicular screw placement in C6 or higher, in order to prevent injuries of the vertebral artery.