Occult Intraspinal Anomalies in Association with An Isolated Congenital Hemivertebra: The Role of Routine Magnetic Resonance Imaging

INTRODUCTION: Previous studies have reported the high incidence of intraspinal anomalies in congenital scoliosis; however, various authors do not consider the presence of an isolated hemivertebra sufficient to warrant magnetic resonance imaging (MRI). To our knowledge, the incidence of intraspinal anomalies detected by MRI and the need for subsequent neurosurgical intervention comparing patients with a single hemivertebra versus patients with a complex pattern involving either hemivertebrae at multiple levels or a hemivertebra coupled with a failure of segmentation is unknown.

METHODS: A retrospective review over a ten-year period (1988-1998) of physical examination findings and plain radiographs in patients presenting with congenital scoliosis and hemivertebrae was conducted to correlate these results with the presence of intraspinal anomalies as detected by MRI and the need for subsequent neurosurgical intervention.

RESULTS: One hundred and sixteen patients with congenital scoliosis and a curve that included at least one hemivertebra were identified. Seventy-six of which presented after 1988 and had an MRI. These patients were the subject of this study and had a mean age at presentation of 59 months (range, 1-198 months) with mean follow-up of 7.1 years (range, 1-17 years). Twenty-nine had an isolated hemivertebra while 47 had a complex hemivertebral pattern. Six patients (21%) with isolated hemivertebra compared to eleven patients (23%) with a complex hemivertebral pattern had an MRI-detected intraspinal anomaly (p=0.99). A positive history or physical exam finding was present in 41% of all patients and was of limited benefit in predicting the presence of intraspinal anomaly in either isolated hemivertebrae (sensitivity 67%, specificity 43%) or complex hemivertebral patterns (sensitivity 55%, specificity 78%). Three patients with an isolated hemivertebra (10%) versus 4 patients with a complex hemivertebral pattern (9%) underwent neurosurgical intervention (p=0.88). All seven patients who underwent neurosurgical intervention had a MRI-detected intraspinal anomalywhile five (71%, 2 complex/3 isolated) hadeither an abnormal history or physical exam finding.

DISCUSSION AND CONCLUSIONS: Isolated hemivertebra and complex hemivertebral patterns have similar incidences of MRIdetected intraspinal anomalies, and subsequent need for neurosurgical intervention. History and physical examination findings were of limited benefit in predicting the presence of MRI-detected intraspinal anomalies and subsequent need for neurosurgical intervention. Therefore, we recommend obtaining an MRI for all cases of congenital scoliosis, including isolated hemivertebrae.