3D- BASED NAVIGATION AT INTERVENTIONS OF THE CERVICAL AND THORACIC SPINE- WHAT IS THE BENEFIT, WHAT ARE THE PROBLEMS?

Abstract

Introduction: In recent years, navigated surgical procedures in spinal surgery have been established due to an increasing demand for precision. Especially 3D-C-arms connected with navigation systems are being used more often and can be utilized intraoperatively for planning as well as controlling of screw positions. This study analyses our experiences with 3D-based navigation in the posterior cervical and high thoracic spine.

Methods: A 3D-C-Arm (Vision Vario 3D, Ziehm) was connected with a navigation system (Vector vision, Brainlab) and since 10/2007 used for the placement of overall 350 Screws at 51 Patients. Of those 9 Patients had to undergo operations in the posterior cervical spine, of 53 screws Judet- (n=8), Massa lateralis- (n=27) and pedicle-screws (n=18) were placed. Indications for instrumentation were traumatic fractures (n=3), spon-dylodiscitis (n=1), multiple metastases with high-grade instability (n=4), and degenerative rheumatic stenosis of the spinal canal (n=1). Concerning the high thoracic spine (T1–10) 42 interventions were made with the method, 297 pedicle- screws were implanted. Indications in this area were traumatic fractures (n=24), metastases (n=14) and spondylodiscites (n=4).

Results: Scan-time intraoperatively took 60 seconds on average, data-transfer to the navigation-system another 10 seconds. Application-time including anti-collision-check needs approx. 6 minutes [5;18]. In total 260/350 (74%) screws could be inserted assisted with navigation, 194/350 (55%) were controlled intraop-eratively. Regarding the cervical spine in 44/53 (89 %) of the screws the navigation procedure was uneventful. Positioning of 37/53 (70%) of the screws was checked immediately postoperatively. In the upper thoracic spine 216/297 (73%) could be placed with navigation, 157/297 (53%) were controlled intraoperatively. Occasionally, scan-setup was problematic, in addition, we experienced technical problems. Correct placement was seen for each screw, thus correlating well with the intraoperative findings.

Conclusions: The application of the combination of intraoperative 3D-imaging and navigation for posterior instrumentation of the cervical and the upper thoracic spine is technically feasible and reliable in clinical use. User- and software-dependant sources of error could be solved during the first course of the series. Image-quality at the cervical spine is depending on individual bone density, and possible metal artifacts. With undisturbed visibility of the vertebral body, the reliability of 3D-based navigation at the cervical spine is comparable to that of CT-based procedures. Additionally, it has the advantage of skipping preoperative acquisition of data as well as the matching-process. Furthermore, exposure to radiation is reduced due to the possibility of sparing pre- and postoperative CT.