Low Sensitivity of Neurological Examination, Curve Laterality, and Thoracic Kyphosis for Detecting Chiari I Malformation and Syrinx in Operative Pediatric Scoliosis

Background: For pediatric patients with presumed adolescent idiopathic scoliosis, surgeons rely on abnormal neurological exam, left-sided thoracic curve, or thoracic hyperkyphosis to initiate preoperative magnetic resonance imaging; however, the sensitivity of these screening criteria for detecting neural axis anomalies of Chiari malformation type I (CM1) and syrinx is not well understood. We aim to determine the frequency of abnormal neurological exam, left-sided thoracic curve, and hyperkyphosis of thoracic curves in patients with operative scoliosis and CM1 and/or syrinx. Methods: A retrospective analysis was conducted with patients who underwent surgical management for pediatric scoliosis and were diagnosed with either CM1 (n=4), syrinx (n=23), or both conditions (n=20). Neurological exams before surgery were reviewed and screened for abnormal deep tendon reflexes, abnormal abdominal reflexes, motor disturbance, sensory disturbance, muscle mass asymmetry, spasticity/contracture, abnormal gait, abnormal balance, and ankle clonus. Thoracic curvature direction, T2-T5 kyphosis, and T5-T12 kyphosis were also obtained. Results: Across all patients with CM1 and/or syrinx that underwent surgery for scoliosis, 9 of 47 (19%) had an abnormal preoperative neurological exam, 11 of 43 (26%) had a left-sided thoracic curve, 5 of 41 (12%) had a T2-T5 hyperkyphosis, and 9 of 41 (22%) had a T5-T12 hyperkyphosis. The T1-T12 kyphosis across the whole cohort averaged 38 degrees. In patients with known CM1 and/or syrinx, the sensitivity of neurological exam abnormality, left-sided curve, or hyperkyphosis was 27 of 41 (66%). Patients with both CM1 and syrinx were not more likely to have an abnormality on imaging or physical exam compared to patients with isolated CM1 or syrinx conditions. Conclusions: These results indicate that neurological exam abnormalities, left-sided thoracic curve, and thoracic hyperkyphosis have a relatively low sensitivity when used as a screening metric for detecting CM1 and/or syrinx in patients with scoliosis. Thirty-four percent of patients with CM1/syrinx and operative scoliosis had a normal neurological exam, right-sided curve, and a normal amount of kyphosis. Surgeons should consider the sensitivity of these screening metrics when determining which patients should obtain a preoperative MRI before scoliosis surgery. Levels of Evidence: Diagnostic studies—investigating a diagnostic test level III with a retrospective analysis.