Curve progression following selective and nonselective spinal fusion for adolescent idiopathic scoliosis: are selective fusions stable?

Purpose: The purpose of this study is to compare postoperative outcomes between selective and non-selective fusions longitudinally over the first five postoperative years. Methods: Patient parameters were retrieved from a multicenter, prospective, database. Patients with Lenke 1–6, B and C deformities were included. Patients were stratified into 2 groups: selective fusion (SF), if the last instrumented vertebra (LIV) was at or cranial to the lumbar apex, or non-selective fusion (NSF). Differences in coronal and sagittal radiographic outcomes were assessed with generalized linear models (GLMs) at 1-, 2- and 5- year postoperative outcomes. Five-year postoperative categorical radiographic outcomes, flexibility, scoliosis research society scores (SRS), and reoperation rates were compared between groups. Matched cohorts were created for subgroup analysis. Results: 416 (SF:261, NF:155) patients, including 353 females were included in this study. The mean preoperative thoracic and lumbar Cobb angles were 57.3 ± 8.9 and 45.3 ± 8.0, respectively. GLMs demonstrated greater postoperative coronal deformity in the SF group (p< 0.01); however, the difference between groups did not change overtime (p> 0.05) indicating a relatively stable postoperative deformity correction. The SF group had a greater incidence of lumbar Cobb ≥ 26 degrees (p< 0.01). The NSF group demonstrated worse forward and lateral flexibility at 5-year postoperative outcome (p< 0.05). There was no difference in postoperative SRS scores between the SF and NSF groups. Reoperation rates were similar between groups. Conclusion: Selective fusion results in greater coronal plane deformity; however, this deformity does not progress significantly over time compared to non-selective fusion. Selective spinal fusion may be a beneficial option for a larger subset of patients than previously identified. Level of evidence: III.