P12. Concave side of proximal thoracic zone vulnerable to pedicle screw perforation in adolescent idiopathic scoliosis surgery: Comparative analysis of pre- and intraoperative computed tomography navigation

Background Context Several reports revealed that various assistive methods, including intra- and preoperative computed tomography (CT) navigation, improve pedicle screw (PS) placement accuracy compared to the conventional freehand technique. However, few studies have clarified the advantage of intraoperative CT navigation, including the second 3D scan, compared with preoperative CT navigation in adolescent idiopathic scoliosisAIS surgery. Purpose To evaluate the accuracy of the PS and clarify the characteristics of pre- or intraoperative perforation using CT navigation in AIS surgery. Study Design/Setting This retrospective radiological study was based on a prospective patient database from a single academic spinal surgery department. Patient Sample One hundred and seven patients who underwent surgery were retrospectively selected. Outcome Measures We analyzed the postoperative CT images for pedicle perforation using Rampersaud criteria. Methods Overall, 853 PSs were implanted in 48 patients using preoperative CT, whereas 1,059 PSs were implanted in 59 patients using intraoperative CT with a second intraoperative 3-dimensional (3D) scan. We analyzed the postoperative CT images for pedicle perforation using Rampersaud criteria. According to these criteria, patients were categorized into four grades: A, entirely in the pedicle; B, <2 mm breach; C, 2–4 mm breach; and D, >4 mm breach. Oversized screws touching the medial and lateral cortices were considered part of grade A unless the breach was >2 mm. Based on the fact that screws perforating the canal for up to 2 mm were considered acceptable, grades A and B were categorized as accurate. Results Total PS placement accuracy (grade A, B) was significantly higher in the intraoperative group than the preoperative CT group (97.2% vs 94.8%, respectively; p=.008). Accuracy was significantly higher among Lenke type 1 AIS cases in the intraoperative CT group (97.8% vs 95.1%, respectively; p=.014). Grade D perforation occurred most frequently at the concave side of proximal thoracic (PT) zone, followed by the transitional PT/main thoracic zone in both groups. The second intraoperative 3D scan allowed the re-insertion of 10 screws to prevent grade D perforation in the intraoperative Conclusions Both navigation systems were sufficiently accurate. However, surgeons must be careful at the concave side of the PT zone. Intraoperative CT navigation yielded higher accuracy by using the intraoperative second 3D scan than preoperative navigation. FDA Device/Drug Status This abstract does not discuss or include any applicable devices or drugs.