The utility of superficial abdominal reflex in the initial diagnosis of scoliosis: a retrospective review of clinical characteristics of scoliosis with syringomyelia

With increasing use of magnetic resonance imaging (MRI), underlying syringomyelia is increasingly found in patients with presumed idiopathic scoliosis. To determine the indications for MRI in the differential diagnosis of scoliosis, several clinical characteristics of syringomyelia have been reported. Neurological signs, particularly abnormal superficial abdominal reflex (SAR), are important in establishing the initial diagnosis of scoliosis. However, the prevalence of abnormal SAR in patients with scoliosis and the sensitivity of this sign in predicting syringomyelia are not well known. We aimed to determine the diagnostic utility of SAR and other characteristics of syringomyelia in patients with scoliosis.

We reviewed the medical records of 93 patients with scoliosis, 90 of whom underwent corrective surgery. All patients underwent MRI to determine the presence of syringomyelia. Mean age at surgery was 12.5 years. Abnormal SAR was defined as unilateral or bilateral absence or hyporeflexia of SAR. We calculated indices of diagnostic utility of abnormal SAR for non-idiopathic scoliosis and for syringomyelia. Abnormal SAR, left thoracic curve pattern, gender, and curve flexibility were compared between scoliosis with syringomyelia and idiopathic scoliosis. Logistic regression analysis was performed with the existence of syringomyelia as the dependent variable and curve flexibility as the independent variable.

Abnormal SAR was observed in 20 patients (prevalence 22%). All 6 patients with myopathic scoliosis displayed bilateral absence of SAR. The sensitivity of abnormal SAR for non-idiopathic scoliosis was 38%, with 96% specificity, 90% PPV (positive predictive value), and 60% NPV (negative predictive value). Syringomyelia was identified in 9 of the 93 patients (9.7%); 8 of these had abnormal SAR. The sensitivity of abnormal SAR for syringomyelia in presumed idiopathic scoliosis was 89%, with 95% specificity, 80% PPV, and 98% NPV. Gender, abnormal neurological findings, and curve flexibility differed significantly between patients with syringomyelia and those with idiopathic scoliosis (P < 0.05). In the logistic regression model, the area under the receiver operating characteristic (ROC) curve was 0.79 and the cut-off value of curve flexibility for syringomyelia was 50% (P = 0.08).

Abnormal SAR was a useful indicator not only for syringomyelia, but also for myogenic scoliosis.