Article
Accuracy of robot-assisted versus fluoroscopy guided pedicle screw insertion for metastatic spinal disease: A matched cohort comparison
Search Medline for
Authors
-
Volodymyr Solomiichuk
- Universitätsmedizin Göttingen, Klinik für Neurochirurgie, Göttingen, Deutschland
-
Veit Rohde
- Universitätsmedizin Göttingen, Klinik und Poliklinik für Neurochirurgie, Göttingen, Deutschland
-
Julius Fleischhammer
- Neuroradiologie, Göttingen, Deutschland
-
Jwad Warda
- Universitätsmedizin Göttingen, Göttingen, Deutschland
-
Awad Alaid
- Göttingen, Deutschland
-
Kajetan von Eckardstein
- Klinik für Neurochirurgie, Universitätsmedizin Göttingen, UMG, Göttingen, Deutschland
-
Bawarjan Schatlo
- Universitätsmedizin Göttingen, Göttingen, Deutschland
| Published: | June 9, 2017 |
|---|
Outline
Text
Objective: Studies validating robotic technology in the spine have hitherto focused on degenerative disease and therefore overrepresent lumbosacral instrumentation where pedicles are relatively large. The aim of the present study is to compare the accuracy of the Mazor/SpineAssist system to conventional fluoroscopy-guided pedicle screw placement for metastatic spinal disease.
Methods: Seventy (n=70) patients suffering from metastatic spinal disease requiring instrumentation were included in the study. Robot-guidance was based on preoperative planning and intraoperative fluoroscopic matching. In the “conventional” group, pedicle screws were placed based on anatomical landmarks and antero-posterior and lateral fluoroscopy. The primary outcome measure was accuracy of screw placement as assessed using the Gertzbein-Robbins scale. Screw grades A and B (less than 2mm pedicle breach) on postoperative coronal and axial CT imaging were considered clinically acceptable, while all other grades were considered misplacements.
Results: A total of 406 screws were placed on 204 levels. The upper thoracic spine (Th1-Th6) accounted for 61 operated levels (29.6%), the lower thoracic spine for 35.9% (74/206), while the remaining (71/206) (34.4%) were lumbosacral. In the robot group (n=35, 192 screws), trajectories were graded A or B in 162 instances (84.4%). The misplacement rate was 15.6% (30/192 screws). In the conventional group (n=35, 214 screws), acceptable trajectories made up 83.6% of screws (179/214) with a misplacement rate of 16.3% (35/214). There was no difference in screw accuracy between groups (Chi2, p=0.89). There was no difference in infection rates between the two groups (5 patients with conventional (14.3%) vs. 8 patients (22.8%) with robotic technique). Moreover, radiation exposure did not differ between groups (p>0.05).
Conclusion: Our initial experience with robotic instrumentation of the thoracic spine is encouraging in that results were comparable to the gold-standard conventional technique. Registration and screw placement were not hampered by osteolytic vertebrae or small thoracic pedicles. On the other hand, radiation exposure and the incidence of wound breakdown were comparable between groups, raising the question whether the added cost and effort justifies robotic surgery for metastatic spinal disease.
