Safety and accuracy of robot-assisted versus fluoroscopy-assisted pedicle screw insertion in thoracolumbar spinal surgery

J Neurosurg Spine 30:615–622, 2019

The object of this study was to compare the safety and accuracy of pedicle screw placement using the TiRobot system versus conventional fluoroscopy in thoracolumbar spinal surgery.

METHODS Patients with degenerative or traumatic thoracolumbar spinal disorders requiring spinal instrumentation were randomly assigned to either the TiRobot-assisted group (RG) or the freehand fluoroscopy-assisted group (FG) at a 1:1 ratio. The primary outcome measure was the accuracy of screw placement according to the Gertzbein-Robbins scale; grades A and B (pedicle breach < 2 mm) were considered clinically acceptable. In the RG, discrepancies between the planned and actual screw placements were measured by merging postoperative CT images with the trajectory planning images. Secondary outcome parameters included proximal facet joint violation, duration of surgery, intraoperative blood loss, conversion to freehand approach in the RG, postoperative hospital stay, and radiation exposure.

RESULTS A total of 1116 pedicle screws were implanted in 234 patients (119 in the FG, and 115 in the RG). In the RG, 95.3% of the screws were perfectly positioned (grade A); the remaining screws were graded B (3.4%), C (0.9%), and D (0.4%). In the FG, 86.1% screws were perfectly positioned (grade A); the remaining screws were graded B (7.4%), C (4.6%), D (1.4%), and E (0.5%). The proportion of clinically acceptable screws was significantly greater in the RG than in the FG (p < 0.01). In the RG, the mean deviation was 1.5 ± 0.8 mm for each screw. The most common direction of screw deviation was lateral in the RG and medial in the FG. Two misplaced screws in the FG required revision surgery, whereas no revision was required in the RG. None of the screws in the RG violated the proximal facet joint, whereas 12 screws (2.1%) in the FG violated the proximal facet joint (p < 0.01). The RG had significantly less blood loss (186.0 ± 255.3 ml) than the FG (217.0 ± 174.3 ml; p < 0.05). There were no significant differences between the two groups in terms of surgical time and postoperative hospital stay. The mean cumulative radiation time was 81.5 ± 38.6 seconds in the RG and 71.5 ± 44.2 seconds in the FG (p = 0.07). Surgeon radiation exposure was significantly less in the RG (21.7 ± 11.5 μSv) than in the FG (70.5 ± 42.0 μSv; p < 0.01).

CONCLUSIONS TiRobot-guided pedicle screw placement is safe and useful in thoracolumbar spinal surgery. Clinical trial registration no.: NCT02890043 (clinicaltrials.gov)