Skip to main content
  • Analysis of Possible Predictive Factors of Spinal Cord Abnormalities on Preoperative MRI in Pediatric Patients with Adolescent Idiopathic Scoliosis

    Final Number:

    Angela Wilcox Palmer MD; Scott H Boop B.S.; Eylem Ocal MD; David Bumpass MD; Richard McCarthy MD

    Study Design:

    Subject Category:

    Meeting: Congress of Neurological Surgeons 2018 Annual Meeting

    Introduction: The use of preoperative magnetic resonance imaging (MRI) in the care of adolescent idiopathic scoliosis (AIS) remains controversial. These patients may have neuro-axis anomalies that pose a risk during scoliosis surgery; manipulation of spinal alignment without knowledge of spinal cord anomalies could risk motor or sensory function. The definitive way to uncover these anomalies is by MRI.

    Methods: We sought to determine predictive factors of spinal cord anomalies in children with AIS, especially those that would require evaluation and intervention prior to scoliosis surgery, such as Chiari malformation and tethered cord. We performed a retrospective chart review of 188 patients and collected demographic information, radiological findings on scoliosis x-rays, and objective signs on the patient’s clinical exam. We performed univariate analysis to discover whether any of these variables might predict the presence of spinal cord abnormalities using logistic regression model. The concordance statistics (C-index) was computed for each potential predictor to assess its predictive accuracy.

    Results: Several variables were noted to have a statistically significant (P-value<0.05) association with the finding of spinal cord abnormality on MRI and demonstrated moderate predictive accuracy. One such variable is Risser classification, an indirect measure of skeletal maturity, calculated when scoliosis surgery was deemed appropriate; the C-index for this variable is 0.715. When a minor structural curve is present, the Cobb angle of this secondary curve also has an association with spinal cord anomaly; C-index is 0.631. The measurement of the sagittal plane deformity is also significant with a C-index of 0.649.

    Conclusions: If children with AIS have spinal cord anomaly, scoliosis deformity corrective surgery could result in a devastating outcome. We describe several variables from scoliosis x-rays which exhibit association with spinal cord anomaly. Further study with more patients may define a predictive model such that spinal cord anomalies could be anticipated without preoperative MRI.

    Patient Care: More and more neurosurgeons are taking care of adolescent idiopathic scoliosis. There are many centers that do not routinely perform MRI, and although the incidence is low of neuroaxis anomalies, such abnormalities as Chiari malformation or tethered cord could pose a serious risk to the patient if they undergo a scoliosis corrective surgery with derotation of the spine. We sought to find predictive variables using the X-rays routinely obtained of these patients as well as demographic and clinical variables to see if there is a way to predict which patients will have these neuroaxis abnormalities. In centers where it is not feasible to obtain MRI on all children, this could direct which children need to obtain an MRI prior to scoliosis surgery to prevent possible injury to the spinal cord or loss of neuromonitoring during these surgeries. Our goal in the care of patients is to keep them free from injury, and we seek to discover a predictive model such that we could direct which patients need to obtain an MRI as they have a high probability of having a neuroaxis anomaly.

    Learning Objectives: By the conclusion of this session, participants should be able to 1) Define adolescent idiopathic scoliosis (AIS) and describe surgical intervention for scoliosis. 2) Understand the effect Chiari malformation and tethered cord have on scoliosis corrective surgical planning and thus why it is important to know whether AIS patients have these abnormalities prior to scoliosis surgery. 3) Describe the clinical, demographic, and x-ray variables that may predict the presence of spinal cord abnormalities and how this can be used in clinical practice.


We use cookies to improve the performance of our site, to analyze the traffic to our site, and to personalize your experience of the site. You can control cookies through your browser settings. Please find more information on the cookies used on our site. Privacy Policy