Right thoracic curves in presumed adolescent idiopathic scoliosis: which clinical and radiographic findings correlate with a preoperative abnormal magnetic resonance image?

Abstract

Study design: Prospective case control study.

Objective: This study investigated preoperative presumed adolescent idiopathic scoliosis (AIS) patients with right thoracic curves to determine which clinical and radiographic findings correlate with neural axis abnormalities on magnetic resonance imaging (MRI).

Summary of background data: Neural axis abnormalities on MRI are not uncommon in patients with left thoracic curves, increased thoracic kyphosis, and in children less than 10 years old. For adolescents with right thoracic curves, less is known regarding which clinical and/or radiographic characteristics accompany neural axis abnormalities.

Methods: A total of 529 presumed AIS patients with Lenke 1-4 right thoracic curve patterns had MRI evaluation before surgery. Thirty-six of these patients had abnormal MRIs (syrinx, Chiari malformation, and/or tethered cord). To differentiate between those with normal MRIs (n = 493) and those with abnormal MRIs (n = 36), the following preoperative clinical parameters were evaluated: age, gender, height, weight, asymmetric abdominal reflexes, thoracic rotation (scoliometer), coronal balance, trunk shift, shoulder elevation, and the Scoliosis Research Society (SRS)-30 questionnaire. Radiographically, thoracic curve magnitude, thoracic rotation (Nash-Moe), coronal balance, trunk shift, length of thoracic curve, location of curve apex, sagittal balance, thoracic kyphosis (T2-T12), and lumbar lordosis were evaluated.

Results: Neural axis abnormalities were found in 6.8%. Those with abnormal MRI findings had significantly greater clinical thoracic rotation (mean difference, 2.4°) and significantly greater radiographic thoracic kyphosis (mean difference 5.9°). However, there were no significant differences in: age (14.9 vs. 14.7 years), height for age (when adjusted for gender), asymmetric abdominal reflexes (3.5% normal MRI group vs. 6.1% abnormal group), coronal balance (clinical or radiographic), trunk shift(clinical or radiographic), shoulder elevation, thoracic curve magnitude (61.4° normal MRI group vs. 63.6° abnormal group), length of thoracic curves (7.0 segments normal group vs. 7.2 segments abnormal group), location of curve apexes, radiographic sagittal balance, or any domains of the preoperative SRS-30 questionnaire.

Conclusion: Of preoperative presumed AIS patients with right thoracic curves who underwent MRI evaluation of the neural axis, 6.8% were found to have neural axis abnormalities, with those having increased rotation and/or increased kyphosis at higher risk. Surgeons should use this information when deciding whether a preoperative MRI is indicated in those with right thoracic AIS curve patterns.