272 Comparative Analysis of the Pedicle Screw Placement Accuracy, Revision Rate and Screw Loosening in a Single Center Using Fluoroscopy, iCT-Based Navigation and Robotic-Guided Surgery

INTRODUCTION: Pedicle-screw (PS)-instrumentation is a common procedure for the management of degenerative, infectious, oncological spine-diseases with instability and spine-fractures. Free-hand and fluoroscopy-guided PS-placement are traditional methods, whereas navigation-guided PS-placement, either with surface-based matching or automatic registration with intraoperative imaging, additionally robotic-guided PS-placement have emerged in the past two decades. Navigation- and robotic-guided surgery shows increased PS-accuracy-placement. METHODS: This retrospective single-center study includes consecutive cases who underwent PS-placement in one of the three techniques by two surgeons. The placement of each individual PS was evaluated in either iCT or postoperative spine-CT, and subsequently classified using Gertzbein and Robbins-System (GRS), Zadichevsky-classification and Heary-classification. Follow-up post-operative CT-scans were performed at a mean period of 8 months following surgery (IQR=5-17) for routine control and fusion assessment. RESULTS: A total of 237 patients (52.3% men) with a median age of 71 years (IQR:61-79) were included. The most common indication were degenerative pathologies (26.2%) and fractures (25.3%). A total of 1352 PS were placed: 667 (49.3%) with fluoroscopy, 241 (17.8%) with iCT-navigation and 444 (32.8%) with robotic-guidance. 12(0.9%) were cervical PS, 640(47.3%) were thoracal PS, 634(46.9%) were lumbar PS and 66(4.9%) were sacral screws. The GRS (p<0.001), Zadichevsky (p=0.002), and Heary-classifications (p<0.001) differed between techniques. Highest accuracy was in robotic-guided group (GRS A=91.7%), followed by fluoroscopy-assisted (GRS A=86.2%), and iCT-navigated placement (GRS A=80.5%). Postoperative revisions due to misplacement (1.6%, P<0.001) or screw-loosening (2.2%, P=0.001) were associated significantly with iCT-navigated-placement compared to two other methods. Screw-diameter =6.5 mm (AUC=0.608; p<0.001) and pedicle-diameter =12 mm (AUC=0.612; p<0.001) predicted long-term screw-loosening. CONCLUSIONS: Robot-guided-placement of PS has shown higher accuracy, less loosening and decreased revision-rate compared to fluoroscopy-guided or iCT-navigated technique. Morphometric characteristics of screw and pedicle-diameter showed predictive value for long-term screw-loosening.