The patient had surgery using a posterior approach (Fig 5). The posterior arch of the atlas was identified and resected (laminectomy of the atlas). The neural canal also was narrowed superiorly by bony elements, necessitating removal extending from the foramen magnum to the lower part of the posterior fossa (suboccipital craniectomy of the foramen magnum) (Fig 5A). The posterior atlantooccipital membrane and ligamentum flavum between the atlas and axis was thickened and excised to achieve complete decompression. We then rotated the atlantoaxial joint into a neutral position and performed a posterior occiput to axis arthrodesis with a right-sided atlantoaxial transarticular screw, wire, and autogenous corticocancellous bone graft (Fig 5B-C). Halo external immobilization was applied postoperatively. Our rationale for the arthrodesis was to stabilize the torticollis. In retrospect, we think this also would have helped to decrease the pseudotumor pannus from the C1 to C2 arthrosis. Although the atlantoaxial joint was transfixed in a neutral position by the transarticular screw, postoperatively, the torticollis was improved but not fully corrected. We presumed that this was related to the spondylosis of the subaxial spine.
At the 6-week followup, the patient was able to walk with the aid of a cane (Fig 6). Power was graded as Medical Research Council Grade 4/5 in the muscles in the lower extremity and as normal in the other muscle groups. The deep tendon reflexes still were increased, but Hoffman's signs, Babinski signs, and ankle clonus were absent. Bladder function had returned to normal. The patient died 2 months postoperatively from cardiac causes.
Only 11 cases of cervical myelopathy from atlantal hypoplasia have been reported,2,6,9-11,13,14,16 the majority in the radiographic and neurosurgical literature. Sawada et al14 described the first case of cervical myelopathy from atlantal hypoplasia. The sagittal diameter of the canal at the level of the atlas was only 7 mm.14 The sagittal diameter of the canal at the level of the atlas in the other reported cases of atlantal hypoplasia provoking cervical myelopathy were from 7 to 12 mm.9,13,16 The normal sagittal diameter of the spinal canal ranges from 16 to 25 mm at the atlas. When the canal is less than 14 mm, spinal cord compression can be suspected.4 Clinical signs and symptoms usually occur when the canal is less than 10 mm.8,20
As in the other reported cases,9-11,13 our patient's symptoms were not present until the seventh decade of life. If one assumes that atlantal hypoplasia is present at skeletal maturity, the late presentations suggest that the hypoplastic atlas does not cause any impairment unless accompanied by another source of spinal canal narrowing such as degenerative changes, trauma, or torticollis. Degenerative changes such as stiffening of the lower cervical spine might increase motion at the upper cervical spine with normal head movements. Radiographic imaging revealed degenerative changes at the lower cervical spine in our patient and in three other patients with atlantal hypoplasia provoking cervical myelopathy.13,16 However, this is not surprising because the patients presented later in life. Had it not been for the cord signal change that focused our attention to the atlantoaxial level, we might have attributed our patient's symptoms of myelopathy to the stenosis at the lower cervical levels. Our patient's symptoms improved with treatment of the upper cervical spine alone.
Our patient also presented with torticollis, which may result from various conditions.5,12,18 It is helpful to subdivide the causes of torticollis into two groups: (1) disorders of the atlantoaxial joint; and (2) other disorders causing limited rotation of the neck without primarily involving the atlantoaxial joint. In the second group, contraction of the sternocleidomastoid muscle results in torticollis. Abnormalities of the internal jugular veins, including jugular vein thrombosis and vertebrojugular fistulae also may cause torticollis.5,12,18 Our patient had enlargement of the left internal jugular vein at the levels of the C4 to C6 vertebrae, although its clinical impact on the torticollis is unknown. We do not think the torticollis was primarily from a malaligned atlantoaxial joint, as correction of this malalignment and arthrodesis did not fully correct his condition. We think severe subaxial spondylosis may have been the primary etiology.
Torticollis narrows the spinal canal at the atlantoaxial level from movement of the ipsilateral lateral mass posteriorly.7 At the extreme of physiologic axial rotation, the spinal canal is reduced to 61% of its cross-sectional area in neutral rotation.17 The axial MR view at the atlantoaxial level (Fig 3B) does not show substantial cord compression. However, the images were obtained with the patient in the supine position with his head rotated to as neutral a position as he could manage, and therefore may not reflect his usual upright neck position. There also was osteoarthrosis in the central atlantoaxial articulation (ie, spur formation at the posterior aspect of the anterior arch of the atlas). The patient had an obvious retrodental pseudotumor from osteoarthrosis of the atlantoaxial joint (Fig 3). Therefore, the late onset of cervical myelopathy may have been attributable to the proliferative changes in the medial atlantoaxial joint caused by degeneration at the C1 to C2 level.
Myelopathy also has been thought to be related to vascular etiologies. Gooding postulated that direct stretching and kinking of arteries occurred during cervical motion and that this could result in intermittent cervical-cord ischemia and loss of vascular autoregulation.3 He suggested that, superimposed on a cord already compromised by external compression, such vascular deficits could lead to a progressive myelopathy.3 During physiologic rotation to the left side, the right vertebral artery becomes somewhat kinked and left stretched because the vertebral arteries running in the transverse foramen are close to the atlantoaxial joint.1,15 Whether the torticollis played a role in our patient's clinical picture through a vascular etiology is conjecture, as dominant vertebral artery kinking typically causes drop attacks rather than myelopathy.3
The current patient had only a 6-week followup. Unfortunately, he died 2 months postoperatively from cardiac causes; therefore we were unable to do longer-term followup. We could be criticized for extending the fusion to the occiput when we might have just fused the atlantoaxial joint. However, at the time of this procedure, the first author (MSP) was not comfortable with decorticating and fusing only the C1 to C2 joint. After resecting the posterior arch of C1, it was our opinion that the most effective means of obtaining solid fusion was extending the arthrodesis to the occiput. If a patient with similar symptoms were to present today, we would limit the fusion to C1 and C2. An MR angiogram, which might have yielded helpful information, was not obtained. In retrospect, we observed that the right foramen transversarium was much smaller than the left (Fig 2B). Although this often is a normal variant in right-handed individuals, it may have been from stenosis and/or occlusion of the right vertebral artery. A preoperative MR angiogram might have answered this question. In addition, it may have provided information regarding the effect of torticollis on the vertebral arteries at the atlantoaxial joint.
Congenital hypoplasia of the atlas is a rare cause of myelopathy. Although the malformation is congenital, the symptoms of myelopathy often do not present until the seventh decade of life, when the myelopathy might be misdiagnosed as age-related physical manifestations. A posterior atlantodental interval less than 14 mm in the absence of atlantoaxial subluxation is consistent with atlantal hypoplasia. Patients with such symptoms should be educated about their condition and observed carefully for signs and symptoms of myelopathy.
1. Barton JW, Margolis MT. Rotational obstructions of the vertebral artery at the atlantoaxial joint. Neuroradiology
2. Fukuzumi Y, Oi S, Matsubara O, Kaito N, Tani S, Abe T. A case of progressive myelopathy with atlas hypoplasia in an infant: controversy in the surgical strategy. Childs Nerv Syst
3. Gooding MR. Pathogenesis of myelopathy in cervical spondylosis. Lancet
4. Greenberg AD. Atlanto-axial dislocations. Brain
5. Kaplan DM, Fliss DM, Peiser Y, Greenberg D, Leiberman A. Internal jugular vein thrombosis in a child due to a ‘pencil point injury’ of the palate. Int J Pediatr Otorhinolaryngol
6. Komatsu Y, Shibata T, Yasuda S, Ono Y, Nose T. Atlas hypoplasia as a cause of high cervical myelopathy: case report. J Neurosurg
7. Mazzara JT, Fielding JW. Effect of C1-C2 rotation on canal size. Clin Orthop Relat Res
8. McRae DL. Bony abnormalities in the region of the foramen magnum: correlation of the anatomic and neurologic findings. Acta Radiol
9. Nishikawa K, Ludwig SC, Colon RJ, Fujimoto Y, Heller JG. Cervical myelopathy and congenital stenosis from hypoplasia of the atlas: report of three cases and literature review. Spine
. 2001;26: E80-E86.
10. Noguchi A, Harada Y, Okabe S, Kohno T, Kamata K, Takahashi H. A surgical case of cervical canal stenosis caused by atlas hypoplasia in an elderly patient. No Shinkei Geka
11. Okamoto K, Sumi M, Ikeda M, Sawamura S, Kataoka O. A case of cervical myelopathy with developmental canal stenosis at the level of the atlas: a case report. Kobe J Med Sci
12. Pearse LA, Sondheimer HM, Washington RL, Robertson D, Clarke DR. Congenital vertebral-jugular fistula in an infant. Pediatr Cardiol
13. Phan N, Marras C, Midha R, Rowed D. Cervical myelopathy caused by hypoplasia of the atlas: two case reports and review of the literature. Neurosurgery
14. Sawada H, Akiguchi I, Fukuyama H, Kameyama M, Koyama T. Marked canal stenosis at the level of the atlas. Neuroradiology
15. Selecki BR. The effects of rotation of the atlas on the axis: experimental work. Med J Aust
16. Tokiyoshi K, Nakagawa H, Kadota T. Spinal canal stenosis at the level of the atlas: case report. Surg Neurol
17. Tucker SK, Taylor BA. Spinal canal capacity in simulated displacements of the atlantoaxial segment: a skeletal study. J Bone Joint Surg Br
18. Vielhauer V, Schewe CK, Schlondorff D. Bilateral thrombosis of the internal jugular veins with spasmodic torticollis in a patient with acquired immunodeficiency syndrome and disseminated cytomegalovirus infection. J Infect
19. Wackenheim A. Radiologic diagnosis of congenital forms, intermittent forms and progressive forms of stenosis of the spinal canal at the level of the atlas. Acta Radiol Diagn (Stockh)
© 2006 Lippincott Williams & Wilkins, Inc.
20. Wolf BS, Khilnani M, Malis L. The sagittal diameter of the bony cervical spinal canal and its significance in cervical spondylosis. J Mt Sinai Hosp N Y