Utilizing Robotic-Assisted Navigation in Early Onset Scoliosis: Experience From a Multicenter Registry

Background: Early onset scoliosis (EOS), defined as scoliosis presenting in patients 9 years of age and younger, presents unique surgical challenges, including dysplastic and narrow pedicles that heighten the risk of pedicle breach and neurovascular injury during scoliosis surgery. Robotic-assisted navigation (RAN) has demonstrated improved pedicle screw placement accuracy compared with conventional freehand and fluoroscopy-guided techniques; however, data on its application in the EOS population remain limited. This study characterizes the intraoperative experiences, technical challenges, and postoperative complication profile associated with RAN in a multicenter EOS cohort. Methods: A prospectively collected, multicenter registry of pediatric patients undergoing spine surgery utilizing RAN across 9 institutions was queried for patients with a diagnosis of EOS treated between January 2022 and April 2025. Demographic, clinical, and surgical characteristics were presented using descriptive statistics. Postoperative complications, including unplanned return to the operating room, infection, and neurological events, were reported in patients with a minimum of 90-day follow-up. Results: Sixty-four patients met the inclusion criteria (mean follow-up: 13 mo, range: 0 to 35); 63% had idiopathic scoliosis, 25% neuromuscular, and 13% other (congenital/syndromic). Mean preoperative curve magnitude was 67±18 degrees. Most patients (88%) underwent a posterior spinal fusion, 11% underwent a growth-friendly surgical procedure, and 2% underwent a revision of a prior growth-friendly construct. A total of 710 robotic screws were placed. Robotic screw success, defined as an attempted screw placed without difficulty or malposition, occurred in 97% of attempted screws overall (710/730), with a mean per-patient success rate of 94%±14%. Malpositioned screws were identified intraoperatively via ball-tipped probe palpation of the pedicle tract and fluoroscopy. Robotic challenges occurred 35 times in 22 patients (34%). Issues with registration (7, 11%) and loss of calibration (9, 14%) were the most frequent challenges encountered. There were 9 malpositioned robotic screws (1% of robotic screws) in 8 patients (13%), all of which were addressed intraoperatively. Nine patients (16%) had a postoperative complication. Two patients had an unplanned return to the operating room, 1 for proximal junctional kyphosis and 1 for deep surgical site infection. Conclusion: This study describes the current perspective associated with the use of RAN in a heterogeneous EOS population, with a high rate of robotic-related technical challenges but a low rate of screw inaccuracy. Level of Evidence: Level III evidence.