Introduction

Atlantoaxial (AA) malformations are a common disease affecting the cervical spine of dogs and often lead to AA instability or subluxation. Clinical signs are often seen at a young age for small-breed dogs but can also occur in medium- and large-breed dogs at any age.^1,2 Atlantoaxial instability has been associated with dorsal spinal cord compression due to suspected dorsal AA ligament hypertrophy, but its clinical relevance and therapeutic approach has not been defined, particularly when dorsal and ventral compression of the spinal cord appears simultaneously.^3,4 Most recently described are the dorsal atlantoaxial compressive bands (ACBs), usually in patients with craniocervical junction abnormalities.^3–10 Although these bands are thought to develop as a consequence of a clinical or subclinical instability affecting the C1–C2 joint, the majority of studies describe them within the context of atlanto-occipital malformations (Chiari-like and atlanto-occipital overlapping).^4–7 Moreover, they are also suggested as an essential component in the development of syringomyelia, present in the greater part of patients with Chiari-like malformations.⁹ It is believed that the original tissue of these bands could be the dura mater or the dorsal AA ligament.^5,8 Histological analysis of biopsied soft tissue has revealed fibrosis with lymphocytic or plasmacytic inflammation and osseous metaplasia.⁷ In some cases, there is evidence of an obvious bony component in the compressive band.⁴ A recent report has described the surgical correction of these bands in three toy-breed dogs with clinical signs due to dorsal compression of the spinal cord but without joint instability.¹⁰

In light of the limited literature describing these lesions and their clinical relevance, this study describes the clinical, radiologic, surgical, and histopathological findings in five dogs with dorsal AA ligament hypertrophy developed as a consequence of suspected joint instability due to a diversity of dens abnormalities and not related with atlanto-occipital malformations, differentiating them from ACBs.

Materials and Methods

Medical records were searched for dogs with dorsal spinal cord compression at the AA junction together with dens abnormalities. Data retrieved included signalment, body weight, neurological status, pre- and postsurgical diagnostic imaging, outcome, and complications.

Results

Five adult or senior dogs (median age: 8.4 yr; range: 3–12 yr) of different breeds (Yorkshire terrier, pug, miniature schnauzer, mixed-breed dog, and cocker spaniel) not usually affected by AA instability (except for a Yorkshire terrier) were referred for assessment of neck pain, associated with tetraparesis in three cases. Mean body weight was 15 kg. None of the dogs had a history of spinal trauma or previous episodes of cervical spinal pain or tetraparesis. The dogs were presented with a variable duration of clinical signs ranging from 1 wk to 2 mo (the median time of signs to presentation at our center was 3.6 wk). Two cases had acute onset clinical signs (less than 1 wk to maximum intensity of clinical signs), which were unresponsive to medical treatment with nonsteroidal anti-inflammatory drugs and restricted exercise. The other three cases were considered chronic and progressive despite the same treatment. The cervical spines of all patients were evaluated radiographically including computed tomography (CT) after intrathecal contrast administration (myelo-CT). Clinical, radiological, and surgical procedures are summarized in Table 1. Radiological examination revealed dorsal spinal cord compression at the AA junction associated with a hypoplastic dens in two dogs (Figure 1A) and a defect in ossification in three dogs that had nonunion of a malformed or fragmented dens (Figures 1B, C). Stress views demonstrated the classic radiological features of AA instability, an increase in the space between the lamina of the atlas and the spinous process of C2, in only two cases (Figure 2). The dorsal compression was considered mild (less than 25% reduction in spinal cord diameter), and these two cases (cases 1 and 2) were treated surgically by means of ventral arthrodesis using transarticular pins and screws and polymethylmethacrylate. Outcome was excellent in both patients, with complete resolution of clinical signs without recurrence 2 yr after intervention.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

TABLE 1 Signalment, Clinical, Radiological, and Surgical Findings in Dogs with AA Instability and Atlantoaxial Dorsal Ligament Hypertrophy

View Fullscreen

Instability was suspected in the remaining three cases because of the presence of a severe dorsal compression (more than 25% reduction in spinal cord diameter) at the level of the C1–C2 joint. In one of these cases (case 3), surgical treatment of dorsal decompression was performed. Postoperative radiographs in forced flexion showed the C1–C2 partial laminectomy without apparent instability (Figure 3A). Nevertheless, 3 days after discharge, the owners reported a sudden vocalization from the patient followed by weakness in all four limbs. An AA subluxation was evident in cervical radiographs with ventral displacement of C2 (Figure 3B). An additional CT was performed, confirming the displacement of the cranial fragment of the nonossified dens and the resultant spinal cord compression (Figure 3C). Surgical fixation of the AA joint was performed by ventral approach to the cervical spine, and the patient recovered uneventfully (Figure 3D). Case 4, in which dorsal compression due to instability was considered severe, was treated in a two-step surgical procedure under a single anesthesia. The procedure included ventral stabilization and dorsal decompression (Figure 4A). During immediate postoperative recovery from general anesthesia, the patient suffered dyspnea attributable to cervical soft tissue swelling. For a duration of 1 wk, the dog received medical treatment and oxygen therapy. Once soft tissue swelling was resolved, the patient recovered normal respiratory airway function without dyspnea and concurrently showed neurological improvement, with only mild tetraparesis remaining 1 yr postoperatively. In cases 3 and 4, dorsal decompressive surgery was performed in the form of a partial C1–C2 laminectomy, removing the caudal aspect of the C1 lamina and cranial aspect of C2 until a band of fibrous tissue, the dorsal AA ligament, was exposed. This tissue was not adhered to the dura mater of the spinal cord. Histopathological analysis revealed a nonvascular, fibrous, noninflammatory connective tissue formed by an amorphous matrix, predominantly noncellular, without evidence of calcium deposits (Figure 4B). These two cases recovered full ambulatory function. Case 3 remained clinically stable without recurrence of clinical signs for 4 yr, when a relapse of an episode of neck pain occurred. A new myelo-CT revealed C6–C7 disc herniation, which was successfully managed with medical treatment (nonsteroidal anti-inflammatory drugs). No instability or compression of the spinal cord at the AA joint was observed. Case 4 was euthanatized for unrelated causes 1 yr after surgery without recurrence of clinical signs. In the last dog (case 5), surgical intervention using the two-step procedure described above was offered to the clients but was declined, and the dog was euthanatized.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Discussion

In this study, all patients showed clinical signs of cervical disease. Radiographically (with the use of myelo-CT), dorsal spinal cord compression at the AA junction was evident in all cases. A relationship between dorsal spinal cord compression caused by ACBs and the presence and severity of syringomyelia in Cavalier King Charles spaniels has recently been described by Cerda-Gonzalez et al.⁹ None of our patients showed obvious syringomyelia or atlanto-occipital abnormalities in CT studies, which could propose an alternative physiopathological mechanism for the disease process than that described for Cavalier King Charles spaniels. There is a similar condition reported by Nakamura et al. in human patients with Chiari I malformation called dural bands, described as a thickening of the dura mater at the atlanto-occipital junction, with degenerative changes and an increase in size and separation of collagen fibers, hyaline nodules, calcifications, and/or ossification.¹¹ Dural bands are considered a cause for syringomyelia owing to the impairment of cerebrospinal fluid flow, so they should be removed (either completely or partially) during the decompressive procedure in patients with Chiari I malformation.^11,12 A similar situation occurs in dogs with Chiari-like malformation; hence, the removal of the dura mater at the atlanto-occipital joint is considered paramount in the decompressive surgery.¹³ Moreover, it has also been described that Chiari-like malformations in dogs can be associated with dural pathology and ligament bands at the AA joint that can indicate potential instability in addition to their occipital dysplasia.^9,10

In our study, the absence of dural abnormalities seen during surgery (cases 3 and 4) suggests that the source of the dorsal compression at C1–C2 is hypertrophy of the dorsal AA ligament. This denomination seems to be more appropriate than the commonly described dorsal ACBs, which is a generic term used to describe a compressive lesion at C1–C2; they are frequently misnamed as dural bands because they have been shown to be caused by fibrosis and proliferation of the ligamentum flavum and dura and are frequently seen in dogs with Chiari-like malformation and syringomyelia.⁵ It also can be hypothesized that these dorsal ACBs are related to subclinical instability caused by an abnormal craniocervical junction. Radiologically, the dorsal AA ligament is better appreciated in sagittal sequences evident as a band, mildly hypointense in T1-weighted and strongly hypointense in T2-weighted sequences.¹⁴ When hypertrophied, it will be presented as a thickened band obliterating the dorsal subarachnoid space or even compressing the spinal cord. As described in other studies, the extended position of the neck during imaging worsened the spinal cord compression in three of our patients (cases 1, 2, and 4).^3,9 This seems reasonable, as neck extension would cause closer proximation of the atlas and axis laminae (attachments of the dorsal AA ligament) and therefore consequent enlargement of the band and the compression. Distinction between dura mater and the dorsal AA ligament by means of diagnostic imaging can be challenging because of the dorsal fusion of dura mater and periosteum between C1 and C2.¹⁵ In humans, and other mammals, including the dog, a vertebrodural ligament that binds dura mater and vertebral canal and a ligament yet to be named projecting from the nuchal ligament between C1 and C2 and tethering itself to the dura mater have been described.^16,17 The large amount of ligamentous structures within this surrounding area makes it difficult to accurately identify the exact anatomical origin of the compressive band.

The authors consider that the hypertrophy of the ligament observed in our cases is caused by a chronic subclinical instability of the AA joint due to the malformed dens. Over time, this instability will cause chronic injury and mechanical irritation, resulting in a compensatory hypertrophy of the ligamentous structures of the joint. Cases 1 and 4 had hypoplastic dens, a common cause for AA subluxation.^1,2 Cases 2 and 5 showed a defect in the proatlas centrum of ossification, and the dens was fragmented in case 5. Case 3 showed a defect in centrum 1, a defect that has been previously described in the dog.^18,19 Interestingly, cases 3, 4, and 5 did not show vertebral displacement in stress radiographs. These cases were the eldest in age and showed greater spinal cord compromise. This suggests that degenerative changes develop slower when there is no evident mobility in stressed position radiographs. Some authors have also suggested that AA instability contributes to spinal cord compression early in life only when it is severe enough to cause subluxation, whereas mild AA instability can result in spinal cord compression at any age.³ The chronic progression of degenerative changes seen in a mildly unstable AA joint may allow the spinal cord to accommodate to the hypertrophy, resulting in delayed onset of clinical signs. From a clinical perspective, it should be noted that the absence of subluxation in stress radiographs does not exclude the possibility of a subclinical instability as demonstrated in case 3, where instability was only evident after dorsal decompression—when luxation of the joint occurred. The authors suggest that evidence of a hypertrophied AA ligament (evidenced as a dorsal compressive lesion at the AA joint) should be considered a direct sign of mild AA instability, particularly in cases with malformed dens.

In human medicine, multiple musculoskeletal conditions such as platybasia, spondylotic spinal changes, and granulomatous inflammatory masses have been described as the consequence of chronic AA instability.^20,21 When affecting the anterior AA ligament, the inflammatory changes may lead to hypertrophy that can be confused with neoplastic lesions, especially in the elderly.²¹ Calcification and ossification of the ligamentum flavum, mainly affecting the thoracic or lumbar spine, are also well described.²² It is believed that degenerative changes found in the ligaments are also a consequence of this continuous and chronic stress.²² This condition has also been described at the AA junction.²³ In such cases, it is frequently impossible to demonstrate an AA instability via radiography, as occurred in our patients.²⁴ Although decompressive surgery is considered the first-choice treatment, many authors have had a good outcome with joint fixation alone.^21,24,25 It is believed that stabilization reverses the degenerative changes caused by instability, as described by Crockard et al.²⁴ This seems to be corroborate with the good outcomes seen in cases 1 and 2 (both having mild dorsal compression), who were only treated with joint fixation. In veterinary literature, stabilization to resolve AA instability in dogs with concurrent dorsal AA ligament hypertrophy has been proved successful.³ A recent paper described a patient with severe dorsal compression associated with AA instability requiring a double approach (dorsal and ventral) for decompression and joint fixation.^4,26 A similar surgical approach was used to resolve cases 3 and 4, where C1–C2 laminectomy and ventral fixation were performed. As a two-step surgery in a single anesthetic procedure may lead to severe postoperative complications (mainly soft tissue swelling and dyspnea), it seems reasonable to plan vertebral fixation initially followed with dorsal decompression some days later. As demonstrated by cases 1 and 2, ventral fixation may be enough to resolve clinical signs. A limitation of the study is that only one dog had MRI of the spine, which is the gold standard for diagnosis of articular diseases and parenchymal spinal cord disease. In the other cases, MRI studies would allow assessment of AA ligaments contributing to stability of the joint and/or syringomyelia.

Conclusion

In conclusion, an abnormal dens may lead to subclinical instability. This subclinical instability may be unnoted in dynamic survey radiographs or imaging techniques but may cause clinical signs in elderly patients as a result of hypertrophic degenerative changes in joint ligaments. Detection of these soft tissue changes at the joint, specifically the dorsal AA ligament, highlights this instability indicating the need for surgical fixation. Such dogs may be treated by ventral fixation alone (if dorsal compression is mild) or in combination with dorsal decompressive procedures. Patients with an abnormal dens and dorsal AA ligament hypertrophy should not go for dorsal decompression without joint fixation because of the risks of luxation. Myelography, myelo-CT, and MRI studies in cervical extension can aid the detection of the severity of dorsal spinal cord compression.