Feasibility and intra-operative accuracy of robotic-guided stacked S1AI and S2AI screw placement.

Abstract

Background: Pelvic fixation with S1 and S2 sacral alar-iliac screws (S1 & S2 AIs) in a stacked configuration can help to offset the exceedingly large forces across the lumbosacral junction in long spinal fusion constructs extending to the sacrum. Traditionally, these modalities of pelvic fixation have been heavily reliant on fluoroscopy and other intraoperative imaging for safe insertion through the sacral alar, the sacroiliac joint and into the narrow corridor of the wing of the ilium. However, recently, computer-assisted robotic guidance and its intraoperative re-registration function has allowed for safe and accurate placement, and intraoperative accuracy assessment of S1 & S2 AIs while minimizing additional radiation associated with fluoroscopic-guidance.

Objective: To review and report our experience with, and assess the intraoperative accuracy and feasibility of, robotic-guided S1AI and S2AI screws inserted in a stacked bedrock configuration as part of instrumented constructs aimed to achieve spinopelvic fusion.

Study design / setting: Retrospective Cohort study.

Outcome measures: We evaluated the number of sacral spinopelvic fixation screws successfully implanted under robotic guidance and the deviation of the achieved screw trajectories from the pre-operatively planned trajectories. The presence and orientation of sacral alar-iliac cortical screw breaches and other intra- or postoperative complications directly related to placement of the stacked S1AI or S2AI implants were also evaluated.

Methods: We included all patients that underwent posterior instrumented spinal fusion up to the pelvis under computer-assisted robotic guidance with open bilateral sacroiliac joint fixation and fusion using S1 & S2 AIs from June 2022 to December 2024. Patients were excluded if the pelvic fixation was not applied in a stacked bedrock configuration. Patient demographics, intra-operative technical errors, complications and other surgical parameters, and any post-operative complications were reviewed and recorded. The data was collected from clinical charts in the electronic medical records system, radiologic images from PACS, and surgical details from operative notes. We also obtained intraoperative secondary registration images for assessment of accuracy, interpreted as the deviation (mm) of the achieved trajectories from the preoperatively planned trajectories in the coronal and sagittal planes, from the robotic planning software system.

Results: 51 patients (32 F & 19 M), mean age 66 (34-80) years, underwent placement of S1AI and S2AI screws under computer-assisted robotic guidance in a stacked bedrock configuration. The most common primary indication for instrumented spinopelvic fusion in these cases was scoliosis (degenerative scoliosis = 26, idiopathic scoliosis = 8), followed in descending order by degenerative spondylolisthesis [10] and kyphosis [7]. Robotic guidance was successful in all 204 screws (4 screws per patient). There was no significant difference in the length of implants inserted on either side of the pelvis, however, the mean length of the S1 AIs (80 mm) was significantly shorter than the S2 AIs (90 mm) [p = 0.005]. The overall mean accuracy in all planes was 2.05 ± 1.97 mm; larger values of proximal deviation were associated with similarly larger values of distal deviation. The mean deviation of the revision screws was greater (3.5 ± 2.16) than that of newly inserted screws (1.72 ± 1.30) [p-value = 0.003]. There were no complications directly related to the robotic-guided placement of the screws. 11 iliac cortical breaches - 5 each through the lateral and medial cortices, and one through the inferior cortex - were identified in 10 patients on post-operative CT images but with no further clinical consequences as they only invaded the bulk masses of the surrounding mm.

Conclusion: Robotic-guided placement of S1 and S2 AI implants in a stacked bedrock configuration for spinopelvic fixation and fusion is feasible with clinically acceptable results. Secondary registration and intra-operative accuracy assessment of screw placement may help to avoid the need for revision surgery related to misplacement of these implants.