Thoracolumbar myelopathies in pug dogs

Abstract Background Constrictive myelopathy (CM) involving a fibrous band around the spinal cord is a newly recognized disease in pug dogs. Objectives To identify the frequency of CM based on diagnostic imaging supplemented with necropsy; to determine whether a relationship exists between the sites of CM and other described T3‐L3 myelopathies; and to determine the frequency of caudal articular process dysplasia (CAPD). Animals Thirty‐two client‐owned pug dogs diagnosed with a chronic, progressive T3‐L3 myelopathy based on neurological examination performed by a board‐certified neurologist. Methods This is a prospective study. All dogs underwent computed tomography (CT) and magnetic resonance imaging (MRI) reviewed by a board‐certified radiologist. Magnetic resonance imaging abnormalities were categorized into diseases; CM only, CM plus other non‐CM condition(s), or non‐CM condition. Sites of CAPD were reported on CT. Nineteen dogs underwent necropsy. Results Magnetic resonance imaging revealed 3 dogs with CM only, 17 with CM plus at least 1 other myelopathy, 11 dogs with non‐CM myelopathies only, and 1 with no MRI abnormalities. Nineteen of 32 dogs had >1 myelopathy diagnosis on MRI whereas 15/32 had >1 site of spinal cord compression. All dogs had CAPD at >1 site in the T3‐L3 vertebral column on CT. Conclusions and Clinical Importance Constrictive myelopathy affected more than half of pug dogs presenting with chronic thoracolumbar myelopathies. Most had multilevel disease, concurrent myelopathies, or both. There was no apparent relationship between anatomic locations of CAPD and most severe myelopathy or myelopathy type.


| INTRODUCTION
Chronic, progressive thoracolumbar (TL) myelopathies are common neurological problems in pug dogs, and multiple abnormalities have been described affecting the TL spinal cord and vertebral column in this breed. 1 Constrictive myelopathy (CM) was first described in pug dogs in 2013 2 and in a 2014 report, pug dogs were cited as the most commonly affected breed with spinal arachnoid diverticulum (SAD) in a study of 122 dogs. 3 Both conditions have been associated with spinal cord compression and meningeal (arachnoid, dural, or both) fibrosis, 4,5 with a constrictive, circumferential band of connective tissue forming in CM and a cerebrospinal fluid-filled, subarachnoid cavity forming in SAD. This could suggest an underlying relationship in the pathogeneses of these conditions. Intervertebral disc herniations (IVDH) are well described in chondrodystrophic and nonchondrodystrophic breeds, 6,7 and caudal articular process dysplasia (CAPD) and hemivertebrae are examples of congenital vertebral column disorders that might not only affect the spinal cord, but which are especially common in pug dogs. [8][9][10] Instability associated with these vertebral malformations has been hypothesized as a reason for development of myelopathy. For example, instability associated with CAPD might lead to CM, which develops gradually over time as a result of chronic micromotion, leading to meningeal fibrosis. 9,11 The fibrous adhesions could also disrupt normal CSF flow, possibly leading to SAD formation. 9 On the other hand, a high percentage (97.0%) of neurologically normal pug dogs in 1 study had CAPD. 12 In another, among 16 pug dogs with CAPD, 5 had no neurological dysfunction. 2 In the 11 dogs with chronic myelopathies, thickened fibrous meninges encompassed the spinal cord in the regions of CAPD, suggesting an association between the vertebral malformation and CM.
A relationship between meningeal thickening by fibrous connective tissue surrounding the spinal cord resulting in neurological dysfunction associated with CAPD was first proposed in 2013 in a condition termed "constrictive myelopathy" which appeared to be unique to pug dogs. 2 Since then, many other names have been used, including "pug myelopathy," "thoracolumbar myelopathy," "meningeal fibrosis," and "thoracolumbar arachnoid fibrosis." 5,10,13 In this report, the authors will use the term constrictive myelopathy to refer to this condition. Despite recent published reports, the pathogenesis, prevalence, and potential relationship of CM with other chronic TL myelopathies such as SAD and IVDH, and with vertebral malformations such as CAPD remain uncertain.
The relatively common occurrence of various types of chronic TL myelopathies in pug dogs, as well as controversy regarding the role of CAPD in these myelopathies, warrant further investigation. Previous studies have not investigated imaging and postmortem findings simultaneously, and many studies excluded conditions such as IVDH or SAD.
The main purpose of this study was to characterize the diagnosis associated with chronic TL myelopathies in pug dogs and to identify the frequency of CM with and without concurrent myelopathies based on diagnostic imaging. Where available, pathological findings are included. A secondary aim was to determine whether a relationship exists between the sites of CM and other described T3-L3 myelopathies (such as IVDH and SAD) and the sites and frequency of CAPD.

| Inclusion criteria
Pug dogs were included in this prospective study if history and neurological examination indicated a chronic (>7 days) progressive T3-L3 myelopathy with no relevant abnormalities revealed by complete blood count and serum biochemistry. A T3-L3 myelopathy was defined as paraparesis of any severity, normal to increased tone and reflexes in the pelvic limbs, with normal thoracic limbs. Consent to postmortem examination was not a requirement for enrollment in this study. Six-month follow-up recheck visits were advised, although if distance from MSU was a hindrance, information was gleaned through contact with dogs' local veterinarians.

| Exclusion criteria
Pug dogs were excluded from the study if neurological examination localized issues to any region of the nervous system other than, or in addition to the T3-L3 spinal cord, or if clinical duration of neurological disease was less than 7 days. Dogs also were excluded if noteworthy non-neurological comorbidities were found.

| Study cohort
Pug dog websites and social media managed by 1 author (K. Smiler) were used for recruitment of most study dogs. Eligibility was determined upon neurological examination and review of bloodwork by 1 author (K. Winger). Signalment and clinical information collected for all dogs included age at onset, characteristics of initial signs and clinical progression, presence of fecal or urinary incontinence, and signs of pain. All dogs underwent diagnostic imaging studies under general anesthesia consisting of dexmedetomidine as a premedication, propofol for induction, and sevoflurane as the inhalant anesthetic. Antisedan was used to reverse the dexmedetomidine, if needed. Dogs that were euthanized or died naturally were submitted for necropsy only with the owner's consent. In studies that included pre-and postcontrast T1-weighted images, the focal dilation of the subarachnoid space was hypointense and noncontrast enhancing, as described. 15 When multiple lesions, sites of compression, or both were present within the same individual, all present myelopathies were noted and that dog was placed in a multiple diagnoses category; in addition, the prominent diagnosis was selected based on the greatest degree of spinal cord disease. Degree of compression was not noted as it was assumed that all compressive lesions could have contributed to the clinical myelopathy. Representative sections of the brain (cerebrum, cerebellum, thalamus, midbrain, pons, and medulla) were also included. If gross lesions were observed, additional sections of brain were submitted for histopathological examination.    Pug dogs were more likely to have multiple concurrent myelopathies than a single myelopathy (Figure 3). Based on MRI findings, there F I G U R E 2 Bar graph representing total number of each myelopathy present in all pug dogs in this study. While caudal articular process dysplasia (CAPD) was present in all studied dogs, constrictive myelopathy (CM) and intervertebral disc herniation (IVDH) were the most frequently encountered compressive myelopathy on neuroimaging. No dogs had CM and subarachnoid diverticulum (SAD) combined, nor SAD alone. F I G U R E 4 Graphical representation of vertebral segments affected by caudal articular process dysplasia (CAPD) (blue) and myelopathies (red). The discrepancy between the CAPD frequency in the mid-to caudal thoracic region, and predominance of non-CAPD myelopathies in the caudal thoracic to cranial lumbar region, makes an association in pathogenesis less likely. Note that the vertebral segments indicate CAPD of the caudal articular process of the cranial listed vertebra, for example, T9-T10 refers to caudal articular process of T9. The caudal-most thoracic vertebra in a few dogs was transitional (with a vestigial rib on 1 side or both sides), we elected to call the caudal-most joint "last thoracic to L1."  Limitations within this study included a lack of neurologically normal control pug dogs, preventing a determination of potential subclinical myelopathies as well as an assessment of CAPD occurrence in normal pug dogs. As CAPD might be present in a majority of pug dogs and our study examined only neurologically abnormal dogs, true prevalence of the condition in the breed remains unknown. Additionally, DNA testing for genetic mutations causing degenerative myelopathy (DM) was not a part of this study; it is possible that DM was yet another diagnosis to consider in these dogs with chronic progressive T3-L3 myelopathies. However, clinical cases of DM in pug dogs are very rare with but 1 report appearing in literature. 17 Importantly, the diagnostic imaging criteria for a CM diagnosis has not been well described at this time; as more is learned about this condition, specific diagnostic criteria may come into sharper focus.

| Diagnostic imaging
This study analyzed the frequencies of common T3-L3 myelopathies in pug dogs, and attempted to correlate vertebral abnormalities and spinal cord diseases. While CAPD was present in 100% of our study dogs, and at several thoracic vertebral sites in each individual animal, the location of the vertebral abnormality was not necessarily associated with the sites of spinal cord disease. This might suggest that CAPD of itself is clinically unremarkable.

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
Approved by the Michigan State University IACUC before enrollment of dogs (protocol 08/14-137-00).

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