Use and Outcome of MRI in the Surgical Treatment of Adolescent Idiopathic Scoliosis

Study Design. Multicenter, prospective, consecutive clinical series. Objective. To report on the use and outcomes of preoperative magnetic resonance imaging (MRI) in a prospective cohort study of 2206 children undergoing posterior spinal fusion and instrumentation for adolescent idiopathic scoliosis. Summary of Background Data. There is no consensus on the use of MRI in the preoperative evaluation of children with idiopathic scoliosis. Also, there is no consensus on the rate of abnormality and the relevance of abnormality on surgical outcome in so-called "idiopathic" scoliosis. Methods. We reviewed the first 2206 patients entered consecutively into the Prospective Pediatric Scoliosis Study, a database of children (8-18 years) undergoing operation for scoliosis by pediatric spinal surgeons in the Spinal Deformity Study Group. Results. There were 1812 girls (80.8%) and 394 boys (17.5%). Mean age at operation was 14 years and 3 months. A total of 191 (8.6%) had juvenile idiopathic scoliosis (age <= 10 years). A total of 923 patients (41.8%) underwent spine MRI. Ninety-one abnormalities of the spine were detected (9.9% of the 923 screened), of which 39 (4.2%) were neural. There were 26 syringes (66.7% of neural abnormalities and 28.6% of all abnormalities), 12 Chiari malformations (30.7% and 13.2%, respectively), and 1 tethered cord (2.6% and 1.1%, respectively). Fifty-three patients (5.7%) demonstrated abnormalities affecting "other" parts of the spine than the neural elements. Patients undergoing MRI more frequently had a thoracic hyperkyphosis (P < 0.001), had a diagnosis of juvenile idiopathic scoliosis (P < 0.001), had a Risser grade between 0 and 2 (P = 0.022), had a greater curve magnitude (P < 0.001), had three major curves (P < 0.001), were male (P = 0.004), and underwent a combined anterior-posterior surgical approach (P < 0.001). Thoracic hyperkyphosis and juvenile onset were associated with greater chance of neural lesion on MRI of the spine. Incidence of abnormal MRI did not differ significantly by direction of apex, Risser grade, curve magnitude or type, male sex, or body mass index. Conclusion. Use of preoperative MRI was 41.8%; 9.9% of patients with so-called "idiopathic" scoliosis had an abnormality on MRI, of which 4.2% were neural anomalies. Of these, syrinx was 66.7%, Chiari was 30.7%, and tethered cord was 2.6%. Significant risk factors for neural abnormality were thoracic hyperkyphosis and juvenile onset. Other characteristics, including apex left thoracic curve, Risser less than or equal to 1, large curve magnitude, triple major curve, male sex, and obesity were not associated with neural abnormality. There were no differences in complication rates between normal and abnormal MRI patients. Our data question the routine use of MRI as a screening tool for adolescent idiopathic scoliosis.