Safe insertion of S-2 alar iliac screws: radiological comparison between 2 insertion points using computed tomography and 3D analysis software.

OBJECTIVE S-2 alar iliac (S2AI) screws are commonly used as anchors for lumbosacral fixation. A serious potential complication of screw insertion is major vascular injury due to anterior or caudal screw deviation. To avoid screw deviation, the pelvic inlet view on intraoperative fluoroscopy images is recommended. However, there has been no detailed investigation of optimal fluoroscopic incline with the pelvic inlet view. The purpose of this study was to investigate the safety margins and to optimize fluoroscopic settings to avoid screw deviation with 2 reported insertion techniques using 3D analysis software and CT. METHODS The study included 50 patients (25 men and 25 women) who underwent abdominal-pelvic CT. With the use of software, the ideal S2AI screws were set from 2 entry points: A) the midpoint between the S-1 dorsal foramen and the S-2 dorsal foramen where they meet the lateral sacral crest, and B) 1 mm inferior and 1 mm lateral to the S-1 dorsal foramen. Anteriorly or caudally deviated screws were defined as deviation of a half thread of the ideal screw by rotation anteriorly or caudally from the entry point. The angular safety margins were compared between the 2 entry points, and patients with small safety margins were investigated. Subsequently, fluoroscopic images were virtualized on ray sum-rendered images. Conditions that provided proper recognition of screw deviation were investigated via lateral and anteroposterior views with the beam tilted caudally. RESULTS The safety margins of S2AI screws were smaller in the anterior direction than in the caudal direction and by entry point A than by entry point B (A: 9.1° ± 1.6° and B: 9.7° ± 1.5° in the anterior direction; A: 10.9° ± 3.8° and B: 13.9° ± 4.1° in the caudal direction). In contrast, patients with a deep-seated L-5 vertebral body tended to have smaller safety margins in the caudal direction. All anteriorly deviated screws were recognized with a 60°-70° inlet view from the S-1 slope. The caudally deviated screws were all recognized on the lateral view, but 31% of screws at entry point A and 21% of screws at entry point B were not recognized on the pelvic inlet view. CONCLUSIONS S2AI screws should be carefully placed to avoid anterior deviation compared with caudal deviation in terms of the safety margin, except in patients with a deep-seated L-5. The difference in safety margins between entry points A and B was negligible. Intraoperative fluoroscopy is recommended with a pelvic inlet view tilted 60°-70° from the S-1 slope to avoid anterior screw deviation. The lateral view is recommended to confirm that the screw is not deviated caudally.