In normal, physiological circumstances there is ample room in the spinal canal to accommodate the spinal cord. Our study aimed to identify the degree of compromise of the spinal canal which could be anticipated in various atlantoaxial pathological states. We examined paired atlas and axis vertebrae using high-definition radiography and simultaneous photography in both normal and simulated pathological orientations in order to measure the resultant dimension of the spinal canal and its percentage occlusion. At the extreme of physiological axial rotation (47°) the spinal canal is reduced to 61% of its cross-sectional area in neutral rotation. The spinal cord is thus safe from compromise. Atlantoaxial subluxation of up to 9 mm reduces the area of the spinal canal, in neutral rotation, to 60% with no cord compromise. Any rotation is, however, likely to cause cord compression. The mechanism of fixation in atlantoaxial rotatory subluxation could be explained by bony interlocking of the facet joint, reproducible in dry bones.
Lateral oblique radiographs are considered important for the identification of spondylolytic lesions, but these projections will give a clear view only when the radiological beam is in the plane of the defect. We studied the variation in orientation of spondylolytic lesions on CT scans of 34 patients with 69 defects. There was a wide variation of angle: only 32% of defects were orientated within 15° of the 45° lateral oblique plane. Lateral oblique radiographs should not be considered as the definitive investigation for spondylolysis. We suggest that CT scans with reverse gantry angle are now more appropriate than oblique radiography for the assessment of spondylolysis. Variation in the angle of the defect may also need consideration when direct repair is being planned.
We reviewed retrospectively the role of monitoring of somatosensory spinal evoked potentials (SSEP) in 99 patients with neuromuscular scoliosis who had had operative correction with Luque-Galveston rods and sublaminar wiring. Our findings showed that SSEP monitoring was useful and that a 50% decrease in the amplitude of the trace optimised both sensitivity and specificity. The detection of true-positive results was higher than in cases of idiopathic scoliosis, but the method was less sensitive and specific and there were more false-negative results. In contrast with the findings in idiopathic scoliosis, recovery of the trace was associated with a 50% to 60% risk of neurological impairment. Only one permanent injury occurred during the use of this technique, and any temporary impairment resolved within two months.
In a prospective trial we performed MRI of the spine and hind brain in 31 patients with scoliosis of onset between the ages of four and 12 years. In eight patients (26%) there was a significant neuroanatomical abnormality; there were six cases of Chiari-1 malformation associated with a syrinx, one isolated Chiari-1 malformation and one astrocytoma of the cervical spine. Four of these patients had left-sided curves. There were no clinical features which could reliably identify those patients with abnormalities on MRI. In particular, the unilateral absence of abdominal reflexes was found to be non-specific (1 of 8 of patients with neuroanatomical abnormalities (12.5%) In view of the established risks of surgical correction of scoliosis in the presence of undecompressed syringomyelia and the possible improvement that may follow decompression of the foramen magnum, we feel that MRI of all patients with scoliosis of juvenile onset should be obligatory.
