Odontoid hypoplasia presenting as torticollis: a discussion of its significance1☆

Case presentation 

A 31-year-old woman presented to a remote out-lying clinic after awakening with painful neck spasms. The patient had no history of trauma or recent illness. At the clinic, the patient received intra-muscular ketorolac and local ultrasound treatment. After she failed to improve with treatment, the attending physician obtained a cervical spine series. The physician interpretation of the X-ray series was C-1 on C-2 subluxation. The patient was then referred to the Emergency Department (ED). The X-ray studies from the clinic did not accompany the patient. She arrived in the ED with a cervical collar, spine board and cervical blocks in place. On arrival, the history and physical examination were repeated. There was no mid-line or bony tenderness and no step-off of a spinous process. She did have palpable left-sided trapezius spasm and tenderness. She had no neurologic deficits and no neurologic symptoms. The patient was maintained in a cervical collar and radiographs were obtained. A cervical series was remarkable for a short stubby odontoid process and no alignment abnormality (Figure 1, Figure 2). In the ED, she received intravenous ketorolac and diazepam. After brief observation and telephone consultation with a neurosurgeon, she was discharged on ibuprofen with the diagnosis of torticollis and congenital hypoplasia of the odontoid process. She was discharged with a soft cervical collar. The following day, the patient was evaluated by the neurosurgeon at a referral hospital (local Korean hospital). The evaluation included a magnetic resonance imaging (MRI) and further radiographs, and she was discharged from care with a report of no evidence of cervical instability. The patient has since been lost to follow-up.

View Large Image Download to PowerPoint

Figure 1. Open mouth view of odontoid hypoplasia.

View Large Image Download to PowerPoint

Figure 2. Lateral cervical spine of the same patient with normal vertebral alignment.

Discussion 

Although their frequency is unknown, congenital anomalies of the odontoid are becoming increasingly more recognized. In congenital conditions such as Down syndrome, Klippel-Feil syndrome and skeletal dysplasia, these anomalies are more common 1, 2. Most are discovered on radiographs after head and neck trauma or spontaneous local symptoms 1, 3. The most common variation is os odontoideum, a round or oval ossicle of bone with smooth borders, which resembles a normal odontoid but is not attached to the body of the axis. Os odontoideum may be in the position of the normal odontoid tip (orthotropic) or less commonly located near the foramen magnum (dystrophic). The second type of odontoid dysplasia is hypoplasia (Figure 1); a variant length of a short odontoid peg exists. The most rare form, and possibly never truly reported, is aplasia of the odontoid or true absence 1, 3, 4, 5. The largest series of os odontoideum have been 9 by Wollin in 1963 and 35 by Fielding in 1980 6, 7. The largest series of odontoid hypoplasia is four from the University Neurosurgical Clinic in Cologne (8).

The vertebral bodies develop by endochondral bone formation from sclerotomal elements derived from the notochord. The axis and atlas are developed from the first and second cervical sclerotomes with contribution from the fourth-occipital sclerotome. The odontoid process is derived from the first cervical sclerotome and is actually the centrum of the body of the atlas, which separates and fuses with the axis 1, 5, 9. The odontoid process ossifies before birth from two ossification centers (9). At birth, the odontoid process is separated from the axis by a cartilaginous band (neurocentral synchondrosis) that represents the vestigial disc space. This band resembles an epiphyseal growth plate on lateral radiographs. It is not at the anatomical base of the dens (the level of the superior articular facets of the axis), but below this level within the body of the axis (5). In this aspect, the odontoid has been described as fitting like a cork in the bottle (body of axis) on an open mouth radiographic view 1, 3.

Radiographic views of the craniovertebral junction can be difficult to interpret. The neurocentral synchondrosis is present in nearly all children at age three and is fused in most by age six. It rarely persists into adult life, but if present, it is a lucency below the anatomical base of the dens. If a fracture were to occur, it would usually occur at the base of the dens and not below the base (3). An os odontoideum may be difficult to differentiate from a non-union fracture of the dens. Radiographic clues of a non-union fracture include an irregular gap between the fragments, which may extend into the body of the axis, bone fragments that seem to match up, and no marginal or smooth cortex (4). In adults, the tip of the odontoid process is located at the level of, or no more than 2 mm above, the rostral border of the anterior arch of the atlas. In less than 5% of adults, the top of the odontoid extends 2–3 mm above the C-1 anterior arch (10). In children under 10 years old, the odontoid tip typically falls short of the top of the anterior arch of the atlas 10, 11. This finding is a normal finding in children and does not represent poor ossification or odontoid hypoplasia (11). Also in children, a terminal ossification center at the apex of the dens may be seen and normally fuses by age 12 (10). Occasionally, it persists as the ossiculum terminale and rarely is of clinical significance (5). Fractures of the dens are typically horizontal and a vertical lucency seen within the dens on radiographs should be viewed suspiciously. Only one near vertical fracture of the dens has been reported and an unfused bifid dens is rare. Look for a C-1 arch fracture or congenital C-1 arch anomaly if a vertical lucency is present over the dens 10, 12. A computed tomography (CT) scan may help to differentiate between a C-1 and C-2 abnormality in this case.

Radiographic findings of an os odontoideum include an independent bone with a well-defined cortex separated from the axis. Aplasia of the odontoid is the lack of an odontoid process beginning at the level of the neurocentral synchondrosis. Hypoplasia of the odontoid typically appears as a stubby peg of an odontoid process but may be of variant length. As mentioned above, hypoplasia of the odontoid may be a difficult diagnosis in children under the age of 10. The clinical issue of odontoid anomalies is alantoaxial stability. This stability can be further evaluated using flexion and extension radiographs, or myelography (3). In odontoid dysplasias, the movement of the odontoid or the lack of an odontoid limits interpretation of flexion and extension radiographs. Fielding et al. suggest using a measurement between a line draw superiorly from the posterior border of the body of the axis in comparison to a line drawn inferiorly from the posterior aspect of the anterior arch of the atlas. Motion greater than 3 mm (a change in this measurement on flexion or extension) is considered pathologic; while in their case series, the average movement was 1.03 cm (range, 0.6–1.5 cm) for those patients with instability (6). Fielding et al. recommend that one criterion for fusion would be excessive instability of greater than 10 mm using this measurement (6). Flexion and extension views are to be conducted voluntarily by the patient, particularly if neurologic complaints are present. The role of CT scan or MRI has not been determined for odontoid dysplasias; however, for patients with current neurologic deficits, these modalities would help to further define anatomy and cord compromise.

The etiology of odontoid dysplasia is assumed to be congenital (13). Other theories have been advanced to include cases of os odontoideum developing years after trauma 1, 6. Also, similar radiographic findings have been described after cervical inflammation (10). The etiology of odontoid anomalies may be congenital, acquired or cases of both; however, the relevant question is the stability of the atlantoaxial joint.

The incidence of odontoid dysplasias is unknown because most patients are suspected to be asymptomatic. Patients may present with local neck pain, torticollis, transient neurologic symptoms or permanent neurologic symptoms 1, 2, 3, 4, 13, 14. They may also present with sudden catastrophic neurologic events to include quadriparesis and vertebral artery insufficiency 8, 13, 14, 15. There are reports of a variety of neurologic symptoms of both the anterior and posterior spinal cord but typically do not have cranial nerve findings (1). The most common neurologic complaints are weakness and ataxia; however, sensory loss, dysesthesias, increased deep tendon reflex and loss of proprioception have been reported 1, 3, 6.

In general, if the clinical manifestations are local neck symptoms without neurologic involvement, the prognosis is excellent. For patients with transient neurologic symptoms after trauma, the usual prognosis is total return of function. However, those patients with insidious or progressive neurologic deficits have greater potential for permanent impairment 1, 3. Neurosurgical and orthopaedic literature recommend posterior cervical fusion as treatment; however, authors disagree on when surgery is indicated. Some authors recommend that all discovered patients should receive surgical fusion due to the potential of a devastating neurologic event 1, 8, 11, 13. Other authors state that patients who are asymptomatic, or have local symptoms without neurologic deficit, may be managed conservatively if no instability is present, arguing the potential of poor surgical outcomes 3, 4, 5, 16.

Conclusion 

Congenital anomalies of the odontoid are rare but carry the possibility of serious neurologic impairment. Patients who present with only local symptoms have been followed from 18 months to more than 3 years without complications and many have had symptoms for years before presentation 6, 8. These patients warrant close follow-up with a spine specialist and precautions due to the risk of trauma. Flexion and extension radiographs are a part of their evaluation but the interpretation is best accomplished by the spine specialist, whether done in the ED or the specialist’s clinic. Patients with neurologic deficits should be immobilized and admitted to an ICU setting. Evaluate these patients for potential respiratory insufficiency. If airway intervention is indicated, strict avoidance of cervical extension should be followed. A CT scan or MRI may help to further differentiate the anatomy; treatment with steroids as with a traumatic spinal cord injury should be based on local protocols. Patients with transient neurologic symptoms have a good prognosis, but they already demonstrate the tendency of instability. It is prudent that these patients receive emergent specialist consultation.