Clinical and magnetic resonance imaging characterization of cervical spondylomyelopathy in juvenile dogs

Abstract Background Cervical spondylomyelopathy (CSM) occurs because of compression of the cervical spinal cord, nerve roots, or both, usually affecting young adult to older large and giant breed dogs. Juvenile dogs are affected infrequently. Objective To describe clinical and magnetic resonance imaging (MRI) findings in juvenile dogs (≤ 12 months) with cervical spondylomyelopathy. Animals Twenty CSM‐affected juvenile dogs. Methods Medical and imaging records for juvenile dogs with CSM were reviewed. History and neurologic examination findings were obtained, including follow‐up data. The MRI studies were reviewed for cause and site of spinal cord compression, intervertebral disk protrusion or degeneration, articular process degenerative changes, intervertebral foraminal stenosis, and spinal cord signal changes. Results Mean (median) age at the time of diagnosis was 9.4 (10) months. There were 16 giant breed dogs. Eighteen dogs had a chronic presentation, 18/20 had proprioceptive ataxia, and 9/20 had cervical pain. On MRI, the principal spinal cord compression occurred at C5‐C6, C6‐C7, or both in most dogs; 12/20 dogs had ≥2 sites of spinal cord compression. The cause of compression was articular process proliferation in 8/20 dogs and disk protrusion in 2/20 dogs. Intervertebral disk degeneration was seen in 9/20 dogs. Follow‐up was obtained for 12/20 dogs: 10/12 were managed medically and 2/12 surgically. Conclusions and Clinical Importance Cervical spondylomyelopathy in juvenile dogs was characterized mostly by osseous‐associated spinal cord compression and multiple compressive sites. Almost half of the dogs had intervertebral disk degeneration. Intervertebral disk protrusion was seen in both giant and large breed dogs.

most commonly affected giant breed is the Great Dane, [3][4][5] whereas the most commonly affected large breed is the Doberman Pinscher. 6,7 The pathogenesis of CSM is not entirely understood, but it is thought to involve dynamic and static factors. These factors also are influenced by vertebral canal stenosis, degree of spinal cord compression, movement-associated compressions, shape of the vertebral canal and vertebral bodies, as well as molecular mechanisms. 1,2 Large breed dogs usually develop CSM after 3 years of age because of disk-associated CSM (DA-CSM) with compression secondary to protrusion of intervertebral disks, and often present with neurologic signs at mean and median ages of 7 and 7.1 years, respectively. 8,9 Giant breed dogs more commonly develop osseous-associated CSM (OA-CSM), which usually results from osseous proliferation of the articular processes leading to spinal cord, nerve root compression, or both around a mean age of 3.8 years and median age of 2.5 years, although they can develop signs at a younger age. 3,5,10,11 One of the difficulties in trying to understand the course of CSM in a population of young dogs is the small number of CSM-related studies that include juvenile dogs. [10][11][12] Another problem is that the exact number of juvenile dogs often is not given in these reports; therefore, it is difficult to know which clinical or imaging findings pertain to young dogs. Of all papers reviewed from the last 45 years that reported juvenile dogs in their population, 3,4,6,10,13-26 only 7 specified how many dogs were ≤ 1 year. 13,15,17,18,[21][22][23] The juvenile form of CSM has not been thoroughly investigated in any study.
Our objective was to characterize the clinical and magnetic resonance imaging (MRI) findings in juvenile dogs (up to 12 months of age) with CSM.

| MATERIALS AND METHODS
Medical and imaging records were searched from 2007 to 2018 for dogs ≤12 months of age with a confirmed diagnosis of CSM based on clinical signs and MRI findings. Search terms included: "cervical spondylomyelopathy," "CSM," and "wobbler." Inclusion criteria for the study therefore were the diagnosis on MRI of CSM in dogs ≤12 months of age at the time of MRI examination and medical records with highfield MRI available for review. Exclusion criteria were age > 12 months or incomplete medical records.
Signalment, history, ancillary diagnostic testing, and neurologic examination findings were obtained from patient records, focusing on age, duration of clinical signs (acute when <1 week, chronic when ≥1 week before diagnosis), presence of ataxia, paresis or both, presence of cervical hyperesthesia, as well as treatment option (conservative or surgical) and follow-up.
The neurologic status of the dogs was graded using a grading system from 1 to 5 (adapted from a previously published system), 27 as follows: grade 1, cervical hyperesthesia only; grade 2, mild pelvic limb ataxia or paresis with or without thoracic limb involvement; grade 3, moderate pelvic limb ataxia or paresis with or without thoracic limb involvement; grade 4, marked pelvic limb ataxia or paresis with thoracic limb involvement; and grade 5, tetraparesis or inability to stand or walk without assistance.
Magnetic resonance imaging of the cervical vertebral column was reviewed by a board-certified neurologist (RCdC). The magnetic resonance images were acquired using a high-field (3.0 T or 1.5 T) MRI scanner, with variable protocols, but included at least sagittal T1-weighted (T1W) and T2-weighted (T2W) views of the cervical vertebral column from C2 to T1 and transverse T1W and T2W views of at least the sites of spinal cord compression. A diagnosis of CSM was confirmed by MRI when ≥1 sites of compression of the spinal cord, nerve roots, or both were observed because of intervertebral disk protrusion, ligamentum flavum hypertrophy, osseous proliferation of the articular processes, or thickening of the dorsal lamina, as well as the presence of absolute or relative stenosis of the vertebral canal. 1 The following information was The criteria to determine these changes followed previous studies on MRI and CSM. 4,8,10,11,28 The cause of spinal cord compression was classified as intervertebral disk protrusion, proliferation of the articular processes, hypertrophy of the ligamentum flavum, thickening of the dorsal lamina, or any combination of these. Spinal cord compression was classified as mild (compression <25% of the cross-sectional area of the noncompressed spinal cord immediately cranial and caudal to the compression), moderate (25-50%), or severe (> 50%). 8,11 The main compression site was chosen as the site with the greatest reduction in cross-sectional area and T2-weighted hyperintensity (when present). F I G U R E 2 Parasagittal T2-weighted (A) and T1-weighted (B), transverse T2-weighted (C) and T1-weighted (D) magnetic resonance imaging of a 12-month-old Great Dane with cervical spondylomyelopathy because of osseous proliferation of the articular processes, more prominent on the left (arrows). Also, note the hypertrophy of the ligamentum flavum (asterisk) dorsal laminectomy and the other had ventral stabilization and dorsal laminectomy. Both reportedly improved after the procedures and still were stable 1.5 and 5 years after surgery, respectively. One dog (1/12), classified as grade 4, died after being attacked by another dog 2 years after diagnosis (before death, the dog had been stable for the first year, but had started to worsen intermittently the last year after a fall). At diagnosis, the dog had been prescribed 0.44 mg/kg prednisone q12h for 7 days, then q24h for 7 days, and then q48h until further notice.  Interestingly, we found disk-associated CSM in a Doberman and a Great Dane at 5 and 7 months of age, respectively. Disk-associated CSM typically is seen in large breed dogs with an average age of 7.9 years, 1 and usually does not develop at such a young age. 6,9 Also, Great Danes are much more likely to develop osseous-associated changes rather than disk-associated changes when younger. 3,5 Although only 2 dogs had DA-CSM, it still was surprising to find dogs with primary lesions associated with disk protrusion at such a young age. Very sparse descriptions of this phenomenon were found in the literature. 13 Also of note was the fact that MRI disclosed changes that were consistent with intervertebral disk degeneration in almost half of the dogs (9/20). This occurrence is much younger than what is usually expected in nonchondrodystrophic breeds, which would be >5 years of age. [29][30][31][32] Degenerative changes in the articular processes also were frequent, with 19/20 dogs having a decreased amount or absence of synovial joint fluid, signs of subchondral bone sclerosis, or some irregularity in the articular surface in at least 1 site. These observations seem consistent with the osseous proliferative changes observed. For foraminal stenosis, no association was observed between neurologic grade and severity of foraminal stenosis.

| RESULTS
Breed distribution for the affected dogs consisted of large and giant breeds, with a higher prevalence of giant breed dogs, primarily Great Danes. This finding was not unexpected, considering the higher prevalence of OA-CSM in this breed, which has been associated with an earlier onset of clinical signs. 3,5 Regarding clinical signs on presentation, the most common sign was proprioceptive ataxia, noted in 18 of 20 dogs. Eleven dogs also showed signs of tetraparesis. Most dogs (90%) had a chronic progressive course over a mean of 7.3 weeks. Several owners reported difficulty in recognizing signs of the disease, particularly in distinguishing the normal, clumsy gait of a puppy from an ataxic gait. In a previous report, the mean duration of clinical signs before diagnosis of OA-CSM in Great Danes was 1.9 years, 33 thus it is possible that in this report several young adults diagnosed with CSM actually started developing signs when only a few months old. In agreement with findings previously reported for older dogs, 1 cervical hyperesthesia was observed in 45% of dogs (9/20) but was not the main reason for presentation in any of the dogs.
Regarding the location of the main spinal cord compression, it was consistent with what has been reported previously in the literature, 6,33 with the majority of lesions affecting the caudal cervical vertebral column. More than half (60%) of the dogs had multiple sites of spinal cord compression, which is not uncommon with CSM. 3 It can be hypothesized that juvenile dogs with CSM had a more severe manifestation of CSM, causing the disease to manifest clinical signs earlier, and that is why they were diagnosed with the disease at such young age. Regarding treatment, most dogs were managed medically, with outcomes varying from stable to progressive decline. The 2 surgically treated dogs were stable 1.5 and 5 years after surgery.
Overall, the presenting neurological signs in the 4 dogs that did not receive long-term corticosteroid treatment suggest that a milder presentation was the reason these dogs were stable on follow-up (the dog with the more severe presentation-grade 4-received 3 weeks of prednisone). However, it is difficult to draw conclusions based on the small number of dogs. Dogs with OA-CSM have been reported to survive several years, whether treated surgically or medically, and improvement is reportedly higher with surgery. 3 Our own experience suggests that short-term outcome usually is positive with medical management, whereas long-term outcome is more variable.
Almost half of the dogs had T2W hyperintensity on MRI. This has been suggested to be associated with a worse prognosis, especially if combined with T1W hypointensity, 34,35 which was not observed in any of the 20 dogs. The proportion of dogs with T2W hyperintensity (45%) is lower than previously described. 3,6,10,36 In humans, T2W hyperintensity with T1W isointensity is associated with changes such as edema, gliosis, Wallerian degeneration, demyelination, and some loss of nerve cells. 37 In dogs, the only experimental study performed suggested a predominance of gray matter changes (motor neuron loss and necrosis). 38 However, if these signal changes in the spinal cord parenchyma are associated with chronicity, 39 perhaps these dogs did not have enough time to develop all of the observable lesions on MRI.
One of the limitations of our study was that follow-up was not available long-term for all 20 dogs, but this limitation is expected in a retrospective study ranging over a decade. A limitation related to the MRI examinations was that different protocols were used. However, a highfield MRI was used in all cases, with at least 3 transverse T1-and T2-weighted slices obtained at the 5 cervical intervertebral disk levels (1 transverse slice at the caudal endplate of the cranial vertebra, 1 at the intervertebral disk, and 1 at the cranial endplate of the caudal vertebra).
Another limitation is the relatively small sample size, although the 20 juvenile dogs described here far exceed the previous sporadic reports of CSM in juvenile dogs.
There are only 2 reports with a larger proportion of juvenile dogs confirmed to have CSM with myelography, CT, or MRI. An MRI study on giant breed dogs with CSM had a large number of young dogs, with 61% of dogs being ≤2 years of age, although it was not specified how many of the dogs were ≤ 1 year of age. 4 A radiographic and myelographic study from South Africa, specifically on the Boerboel breed with CSM, found a higher proportion (70%) of juvenile dogs with CSM. 23 It is possible this is a reflection of the large number of Boerboels in that region, and also of the presentation of the disease in that breed.

| CONCLUSIONS
Cervical spondylomyelopathy in juvenile dogs generally was characterized by a chronic presentation, mostly affecting giant breeds, with proprioceptive ataxia with or without tetraparesis as the most common gait abnormality. Compression of the spinal cord occurred most frequently at the caudal cervical vertebral region because of osseous-associated changes. It was not uncommon for dogs to have >1 site of compression.
Compressive lesions caused by intervertebral disk protrusion should not be automatically ruled out because of the young age or breed of a dog, because they were noted in 2 dogs in our study and almost half of the population had intervertebral disk degeneration. Spinal cord parenchyma signal changes were not predominant within the group, and only T2W hyperintensity was observed. Outcome varied greatly, including stable without long-term corticosteroid administration, stable with surgery, waxing and waning of signs or progressive worsening despite long-term corticosteroid administration.

ACKNOWLEDGMENTS
Dr. Bonelli received a scholarship grant from the Coordination for the Improvement of Higher Education Personnel (CAPES), Brazil. CAPES was not involved in any stage of this study.

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
This is a retrospective study. For dogs prospectively enrolled in past studies, all procedures were performed with the approval of the Clinical Research Advisory Committee and the IACUC of The Ohio State University.

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.