Diagnosis and Treatment of a Chronic Atlanto-Occipital Subluxation in a Dog

Introduction

Atlanto-occipital subluxation is rare in dogs and cats, probably because of the stable nature of this joint. When it occurs, central nervous system damage is usually severe enough to cause death, and survival has rarely been reported.^1–5 The stability of the atlanto-occipital joint is attributable to the multiple, strong ligaments associated with it.^2,6 The atlanto-occipital joint is continuous with the atlanto-axial joint by the dens and the associated ligaments. The dens is maintained in the fovea dentis of the atlas by the transverse ligament of the atlas.⁶ The single apical and paired alar ligaments course from the cranial portion of the dens to the basioccipital bone between the occipital condyles.⁶ Additional lateral stability to the atlanto-occipital joint is provided by the paired lateral atlanto-occipital ligaments. These ligaments connect the lateral part of the dorsal arch of the atlas to the jugular process of the occipital bone.⁶ The dorsal atlanto-occipital membrane connects the dorsal edge of the foramen magnum with the cranial border of the dorsal arch of the atlas.⁶ A study of 57 dog cadavers concluded that extreme hyperextension of the atlantooccipital joint was probably the cause of atlanto-occipital dislocation.⁷

The purpose of this report is to describe the myelographic and computed tomographic findings in a dog treated surgically for chronic atlanto-occipital subluxation. Decompression of the brain stem by partial removal of the occipital bone resulted in resolution of the clinical signs.

Case Report

A 6-year-old, 37-kg, castrated male Labrador retriever-cross was referred to the Veterinary Medical Teaching Hospital (VMTH) at the University of Wisconsin, Madison, for evaluation of an abnormal gait and head carriage 6 weeks after having been hit by a car. The referring veterinarian had treated the dog with carprofen^a (2.7 mg/kg per os [PO] q 12 hours) and methocarbamol^b (20 to 27 mg/kg PO q 8 to 12 hours as needed) with resolution of the neck pain, but the dog developed a tetraparesis that was worse in the thoracic limbs. Treatment with prednisone^c (0.5 mg/kg PO q 12 hours for 5 days, q 24 hours for 5 days, and then q 48 hours for 10 days) resulted in significant improvement of the tetraparesis; however, the clinical signs recurred when the prednisone was tapered.

Upon presentation to the VMTH, the dog carried his head and neck stiffly, and palpation of the cervical spine revealed an asymmetrical positioning of the atlas, with the left transverse process being more dorsal than the right. The neck muscles were atrophied. Other neurological abnormalities included ambulatory tetraparesis, pronounced ataxia, and absent conscious proprioception in all limbs. Intermittent dragging of all limbs during ambulation was worse in the thoracic limbs. Mentation as well as cranial nerves and spinal reflexes were normal. No pain was elicited on palpation of the spine. The dog resisted flexion of the head ventrally, and he was stiff on turning his neck to either side. These findings suggested a lesion between the first and fifth cervical (C₁–C₅) spinal cord segments.

A complete blood count, biochemical panel, thoracic radiography, and abdominal ultrasonography were all normal. A urinalysis and urine culture revealed a urinary tract infection with Escherichia coli that was treated with amoxicillin/ clavulanic acid^d (15 mg/kg PO q 12 hours).

Spinal radiography, myelography via a lumbar injection, and computed tomography (CT) were performed with the dog under general anesthesia. Lateral views of the plain radiographs and the myelogram demonstrated a rotation of the C₁ vertebra, with the left transverse process being more dorsal than the right. The caudal fossa and brain stem were deviated ventrally just rostral to the C₁ vertebra. On myelography, a dorsal compression of the spinomedullary junction and brain stem by the occipital bone was noted, with thinning of the dorsal contrast column at the same site [Figure 1A].

Ventrodorsal views of the radiographs and myelogram demonstrated a subluxation of the atlanto-occipital joint, with increased joint space at the right articulation and no visible joint space at the left articulation. The caudal fossa and brain stem were deviated to the left, rostral to the C₁ vertebra, and thinning of the lateral contrast columns was seen on the myelogram [Figure 1B]. Flexed views under fluoroscopy were obtained to rule out an atlantoaxial subluxation. Reduced range of motion and stiffness of the atlanto-occipital joint were noted, and the dens axis and the atlantoaxial joint were normal.

Computed tomography of the caudal brain stem and cervical spine was performed using a third-generation CT scanner. ^e The C₁ vertebra was displaced dorsally and medially at the left atlanto-occipital joint. The left occipital condyle was located outside the cranial articular fovea of the atlas, while the right occipital condyle was located within the spinal canal. The dorsal lamina of the C₁ vertebra was overriding the occipital bone. The skull was subluxated ventrally, resulting in dorsal compression of the brain stem and the spinomedullary junction from the occipital bone [Figures 2A, 2B]. No fractures of the occipital bone or the C₁ and second cervical (C₂) vertebrae were detected, and no evidence of compression of the spinal cord by the C₁ vertebra was seen [Figure 3]. The dens axis appeared normal. A three-dimensional CT reconstruction helped to confirm the absence of fractures of the occipital bone and C₁ and C₂ vertebrae, and it enabled better visualization of the subluxation [Figures 4A, 4B, 4C].

Cerebrospinal fluid obtained prior to myelography via lumbar puncture had an elevated protein content (50.1 mg/dL; reference range <35 mg/dL) and an elevated nucleated cell count (10/μL; reference range <5/μL). On a cytospin preparation, 92% of the cells were lymphocytes and 8% were macrophages. These results were considered nonspecific and were similar to findings previously described with chronic spinal cord compression.^8,9

Two days after the CT, the dog was anesthetized to reduce the subluxation. Fentanyl^f (2.5 to 5 μg/kg per hour intravenously [IV] as a constant-rate infusion [CRI]) was started for pain control. After closed reduction of the atlanto-occipital subluxation under fluoroscopy failed, decompressive surgery was performed. Atracurium besylate^g (8.2 mg/kg IV total) was given in divided boluses during the surgery to paralyze the animal. The dog was placed in sternal recumbency with the neck in a flexed position. A dorsal midline incision over the cranial cervical spine was made. Dissection was done laterally to the transverse processes of the C₁ vertebra to expose the lateral vertebral foramina and the atlantooccipital joints. A large amount of fibrotic scar tissue around the joints and surrounding fasciae prevented reduction of the subluxation. Aggressive removal of the scar tissue was not done in order to avoid instability of the atlanto-occipital joint. The dorsal atlanto-occipital ligament was removed. A craniectomy was done by removing a 1 cm-diameter piece of the occipital bone above the foramen magnum using a high-speed burr. The enlargement of the foramen magnum resulted in good decompression of the medulla oblongata and the spinal cord. Alarge amount of hemorrhage occurred from the right vertebral vein and artery during surgery, which was controlled with application of gel foam^h and bone wax.ⁱ Gel foam was also placed over the part of the cerebellum no longer covered by the occipital bone, to protect it from the epaxial muscles. Closure of the epaxial muscles and skin incision was routine.

Postoperatively, pain was controlled with a CRI of fentanyl^f for 12 hours and application of a transdermal fentanyl patch^j (75 μg). Dexamethasone^k was given immediately postoperatively (0.25 mg/kg IV) and the day after surgery (0.1 mg/kg IV) to reduce inflammation. Amoxicillin/clavulanic acid was continued for the urinary tract infection. Prednisone^c (0.4 mg/kg PO q 12 hours for 5 days, q 24 hours for 4 days, 0.2 mg/kg PO q 24 hours for 3 days) was started 24 hours postoperatively to reduce inflammation.

The day after surgery, the dog was mildly obtunded, non-ambulatory, and tetraparetic, but he could support weight in all limbs when assisted. On day 2 after surgery, a slight head tilt, falling to the left side, and intermittent obtundation were noted, which were interpreted as central vestibular signs. The dog was able to stand but could not walk. At discharge 10 days after surgery, the dog was able to walk a few steps but would try to walk backward instead of forward.

On reevaluation 34 days after surgery, neurological abnormalities included a mild left head tilt and mild tetraparesis, but no dragging of the feet. Conscious proprioception was absent in the thoracic limbs, delayed in the right pelvic limb, and normal in the left pelvic limb. A neurological examination performed 6 months after surgery by the referring veterinarian was normal. By 8 months after surgery, the dog had resumed all normal activity at home.

Discussion

Atlanto-occipital subluxation has rarely been reported in dogs. In previously reported cases of traumatic atlanto-occipital subluxation (in five dogs and one cat), management and resolution of clinical signs relied on open (n=1) or closed (n=4) reduction of the subluxation and stabilization of the joint via either surgical implants (n=1) or external coaptation (n=4).^1–5 One case was treated with a dorsal laminectomy of the C¹ vertebra and interarticular wire and an external splint.⁵ Of the previously reported cases, most were presented after an acute episode of trauma. One dog was presented 12 days after a traumatic incident.⁴ In the case reported here, the dog was presented 6 weeks after the initial trauma, and closed or open reduction of the subluxation was not possible because of the chronic changes present.

In humans, atlanto-occipital subluxation is usually fatal; however, in some cases the clinical signs can be mild or subclinical and cause a delay in the diagnosis.^10,11 The head position of the dog in this case may have occurred because of the subluxation of the left atlanto-occipital joint, or it may have been secondary to pain. In people with chronic atlanto-occipital subluxation, pain is not always present.^10,11

Only in one previous report were CT and three-dimensional CT reconstruction used for treatment planning in dogs.¹ In the case reported here, the survey spinal radiographs and the myelogram helped identify the atlantooccipital subluxation and the location of the compression. The CT and three-dimensional reconstruction images provided better understanding of the subluxation, which helped in planning the appropriate surgical approach. In addition, vertebral fractures, which can easily be overlooked on radiographs, were ruled out with CT.¹²

In one report of a chronic atlanto-occipital subluxation in a human, decompressive surgery was the recommended treatment instead of stabilization of the subluxation if the subluxation could not be reduced.¹³ The neurological signs of the dog in the current case were thought to be secondary to the compression of the brain stem and cervicomedullary junction by the occipital bone, which was later supported by resolution of the clinical signs after decompressive surgery; however, a repeat CT or myelogram was not done to confirm complete decompression of this area. The muscle atrophy in the neck most likely occurred from compression of the spinal roots of the accessory nerve as it enters the foramen magnum.⁶ One prior report in a dog described hypoglossal nerve deficits associated with atlanto-occipital subluxation and resolution of the clinical signs once the subluxation was reduced.¹

In the dog reported here, the neurological signs were worse in the thoracic limbs than in the pelvic limbs. In people, this is a rare finding and has been described primarily with trauma to the cranial cervical spine.^13–15 Theories to explain this phenomenon involve cruciate paralysis or central cord sign.^13–15 With cruciate paralysis, an injury to the cervicomedullary junction results in selective upper-motor neuron dysfunction from damage to the corticospinal decussation. ¹⁴ With central cord sign, an injury in the central part of the spinal cord affects the medially located motor pathways to the thoracic limbs more than the laterally located pathways to the pelvic limbs.¹³ Both phenomena may result from similar pathological mechanisms.¹⁵ Compression of the central region of the cervical spinal cord from a ventral midline site that results in more severe clinical signs in the thoracic limbs has also been recognized in dogs.¹⁶ In the case reported here, the medial part of the cervicomedullary junction was the most severely compressed area.

Immobilization of the neck with an external cast seemed to be crucial for a successful outcome in previously reported cases in animals.^1–5 In the dog reported here, the fibrous scar tissue around the atlanto-occipital joints provided stability, and immobilization of the neck was deemed unnecessary.

The decompressive surgery resulted in resolution of clinical signs in this case. The deterioration in mentation and vestibular signs that occurred after surgery in the dog reported here may have several explanations. First, reperfusion injury affecting vestibular nuclei in the brain stem and spinocerebellar pathways may have occurred after decompression of a longstanding compressive lesion.¹⁷ Second, the bleeding from the vertebral artery that occurred during surgery may have caused ischemia to the brain. In people, vertebral artery occlusion secondary to spinal trauma may cause brain-stem ischemia with secondary neurological signs.^18,19 In people and dogs, the vertebral arteries and basilar artery provide the main arterial blood supply to the most caudal part of the cerebrum, the entire brain stem, and the cerebellum.²⁰ The basilar artery enters the foramen magnum and may have been compromised already in this dog, causing the brain to be vulnerable to an ischemic event.

Conclusion

A 6-year-old dog was diagnosed with an atlanto-occipital subluxation 6 weeks after a traumatic event. Neurological signs resolved after surgical decompression of the brain stem by partial removal of the occipital bone. Computed tomography provided good visualization of the compression of the spinomedullary junction, and it was helpful in ruling out vertebral fractures and planning surgery. Decompression without stabilization may be a viable treatment in chronic cases of atlanto-occipital subluxation if the subluxation cannot be reduced.