Introduction

Intervertebral disk herniation is a common neurologic condition most frequently affecting the thoracolumbar region in dogs. Disk herniations at the cervical level are less common, accounting for approximately 14% to 35% of all herniations.^1–3 Cervical disk herniations, in particular acute extrusion, are reported to cause less severe neurologic deficits than thoracolumbar herniations,^4,5 and this difference has been attributed to the relatively larger size of the vertebral canal in the cervical region, compared with the size of the spinal cord, with the assumption being that there is a greater epidural space at this level.^6–10

Several morphometric studies evaluating the dimensions of the canine vertebral canal and spinal cord have been published. Myelographic,¹¹ CT,¹² and MRI¹³ studies of the cervical spinal cord and vertebral canal ratio have been performed. MRI has additionally been used in the thoracolumbar region.¹⁴ Lim et al¹⁵ used MRI and CT to create reference values for spinal cord and vertebral canal dimensions in clinically normal Beagles and dogs with spinal disorders in the cervical, thoracic, and lumbar regions. They found little difference in spinal cord-to-vertebral canal ratios between the cervical and thoracic regions in a small population of healthy Beagles. An extensive electronic search in PubMed and Google Scholar performed by the authors with the keywords “dog,” “vertebral canal,” “spinal cord,” “ratio,” “measurement,” “diameter,” and “epidural space” did not retrieve any studies specifically comparing the spinal cord-to-vertebral canal ratios in the cervical and thoracolumbar regions in a larger group of dogs.

We have observed that the spinal cord is often severely compressed but with minimal accompanying neurologic deficits following intervertebral disk extrusion in the cervical region and that, conversely, the spinal cord can often seem to not be severely compressed following disk extrusion in the thoracolumbar region, even when marked neurologic deficits are seen. The reason for the often-minimal deficits with cervical intervertebral disk extrusion, versus thoracolumbar extrusion, was therefore hypothesized to be due to something other than a larger epidural space in the cervical region, as is often stated. The objective of the study reported here was to compare the spinal cord-to-vertebral canal area ratios measured with CT between the cervical and thoracolumbar regions in a population of French Bulldogs. We hypothesized that there would not be a significant difference in the ratios between the 2 regions.

Materials and Methods

The digital medical records of the Queen’s Veterinary School Hospital for 2016 through 2019 were reviewed to identify French Bulldogs that underwent CT for assessment of brachycephalic obstructive airway syndrome. Animals were included if complete medical records and imaging studies ranging at least from C5 to L2 were available. Only French Bulldogs ≥ 12 months of age were included. Exclusion criteria were the presence of vertebral malformations or evidence of spinal cord compression at the sites where measurements were obtained and incomplete growth plate fusion. Animals with cervicothoracic or thoracolumbar transitional vertebrae were additionally excluded, as were those with reported neurologic deficits. Dogs for which CT image quality was deemed inadequate to perform the measurements by both observers were also excluded. The medical records were reviewed by a single author to determine study eligibility. Information retrieved from the medical records included age and sex as well as physical and neurologic examination findings. The study was approved by the Ethics and Welfare Committee of the Department of Veterinary Medicine, University of Cambridge (reference No. CR562).

CT was performed with a 16-slice CT scanner (Aquilion 16; Toshiba America Medical Systems). Images were independently reviewed, and measurements were made by 2 blinded assessors (SS and M-AG), one of them a board-certified radiologist (M-AG) and the other a neurology resident in training (SS). Images were viewed with the 2019-2020 Horos Project Digital Imaging and Communications in Medicine viewer. Measurements were made with a lung window reconstructed for bone at the level of the midbody of C5 and midbody of L1 (Figure 1). An area measurement tool was used to trace a closed polygon around the vertebral canal and spinal cord at both sites, as previously described.^12,15 The window width was 1,500 HU, and the window length was 300 HU. For dogs in which a large basivertebral foramen was present, measurements were made immediately cranial to the vertebral midbody. The spinal cord-to-vertebral canal area ratio was calculated by dividing the spinal cord area by the vertebral canal area measured at each site.

Transverse CT images of a healthy French Bulldog obtained at the level of the midbody of C5 (A) and the midbody of L1 (B). The images illustrate the use of an area measurement tool to trace closed polygons around the vertebral canal and spinal cord.

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.06.0266

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Transverse CT images of a healthy French Bulldog obtained at the level of the midbody of C5 (A) and the midbody of L1 (B). The images illustrate the use of an area measurement tool to trace closed polygons around the vertebral canal and spinal cord.

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.06.0266

• Download Figure
• Download figure as PowerPoint slide

Transverse CT images of a healthy French Bulldog obtained at the level of the midbody of C5 (A) and the midbody of L1 (B). The images illustrate the use of an area measurement tool to trace closed polygons around the vertebral canal and spinal cord.

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.06.0266

• Download Figure
• Download figure as PowerPoint slide

Results

The search of the digital medical record database initially identified 54 potential cases admitted between 2016 and 2019. Of these, 17 were further excluded because imaging studies did not extend from C5 to L2 (n = 11), thoracolumbar (3) or cervicothoracic (1) transitional vertebrae were present, neurologic deficits were reported (1), and spinal cord definition on CT images was inadequate (1). The remaining 37 cases were included in the study (Supplementary Table S1).

All of the dogs were French Bulldogs. Twenty-two (59%) were male, and 15 (41%) were female. Median age was 12 months (range, 12 to 84 months).

A Shapiro-Wilk test revealed that data were normally distributed for one of the observers (M-AG) but not for the other (SS; Figure 2). A 2-tailed, paired t test indicated that the mean spinal cord-to-vertebral canal area ratio differed significantly (t = 9.11; df = 36; P < .01) between C5 (0.726) and L1 (0.605) for the first observer (M-AG). A 2-tailed Wilcoxon signed rank test indicated that the median spinal cord-to-vertebral canal area ratio also differed significantly (W = 681; P < .01) between C5 (0.7332) and L1 (0.6132) for the second observer (SS). For both observers individually, the spinal cord-to-vertebral canal area ratio was significantly higher at C5 than L1.

Box-and-whisker plots of spinal cord-to-vertebral canal area ratios measured with CT at the midbody of C5 and midbody of L1 by 2 observers (M-AG [A] and SS [B]).

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.06.0266

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Box-and-whisker plots of spinal cord-to-vertebral canal area ratios measured with CT at the midbody of C5 and midbody of L1 by 2 observers (M-AG [A] and SS [B]).

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.06.0266

• Download Figure
• Download figure as PowerPoint slide

Box-and-whisker plots of spinal cord-to-vertebral canal area ratios measured with CT at the midbody of C5 and midbody of L1 by 2 observers (M-AG [A] and SS [B]).

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.06.0266

• Download Figure
• Download figure as PowerPoint slide

Discussion

Results revealed that the spinal cord-to-vertebral canal area ratio in this group of French Bulldogs was lower in the thoracolumbar than in the cervical region, suggesting that, contrary to popular belief, the epidural space is not larger in the cervical region than the thoracolumbar region, at least in this group of dogs.

Despite the similar pathophysiology between cervical and thoracolumbar intervertebral disk extrusions, they typically result in distinct clinical signs and are associated with differing outcomes,^7,9,16 with thoracolumbar extrusions more commonly leading to clinically important neurologic deficits than cervical extrusions. This has been suggested to be due to the relatively larger vertebral canal size, relative to the size of the spinal cord, in the cervical region versus the thoracolumbar region.^6–10 However, in our experience, it is common to find substantial spinal cord compression associated with intervertebral disk extrusions in both regions, and there is limited evidence of an association between the severity of compression and either neurologic presentation or outcome.^17,18 Results of the present study suggest that, contrary to previous reports, the epidural space is not larger in the cervical region than in the thoracolumbar region in this population of dogs, and an alternative reason for the differing nature of the signs seen associated with intervertebral disk extrusion at each region should be considered.

A population composed exclusively of French Bulldogs without neurologic signs was chosen for the present study given the high number of CT studies available for review at our institution, the increasing popularity of the French Bulldog breed, and the predisposition of French Bulldogs to develop both cervical and thoracolumbar intervertebral disk extrusions.¹⁶ Measurements were made at the midbody level of C5 and L1 because the C5-6 and T13-L1 intervertebral disk spaces are among the most commonly reported sites for intervertebral disk herniation in dogs^16,18,19 and because of the availability of a large number of CT studies spanning that region, the generally uniform appearance of these vertebrae, and the ease of measurement at these sites. Two observers independently reviewed the images and made measurements to demonstrate the repeatability of the findings, albeit under common conditions. Each observer’s measurements were statistically analyzed, and for both, we found that there was a significantly smaller spinal cord-to-vertebral canal area ratio at L1 than at C5, indicating a potentially larger epidural space in the thoracolumbar region and not in the cervical region, as previously reported. Because the objective of our study was simply to make a direct comparison of the spinal cord-to-vertebral canal area ratios in the cervical and thoracolumbar regions and not to provide reference values for these ratios, it was considered more important that each reviewer’s data pointed toward the same conclusion (ie, that the conclusion was repeatable) than whether they agreed on the individual measurements. For this reason, the authors chose to analyze each reviewer’s data independently and not evaluate interobserver agreement or undertake rank analysis. Therefore, future studies are required to corroborate these findings and the values reported should not be used as any form of reference value.

The reason for the often more severe neurologic deficits observed in dogs with thoracolumbar versus cervical intervertebral disk extrusions remains unclear. The premise that a larger epidural space may be a protective finding with intervertebral disk extrusions is undermined by evidence that the degree of spinal cord compression in thoracolumbar disk extrusions is unrelated to the neurologic findings or prognosis.¹⁸ Differences in epidural material and inflammation between the cervical and thoracolumbar regions in dogs with intervertebral disk extrusions have been reported. The cervical region has been associated with less extensive inflammation and less severe calcification.⁹ However, an association between severity of epidural inflammation and neurologic grade has not been found.²⁰ Biochemical, metabolic, anatomic, and biomechanical differences between the 2 locations have been hypothesized to be the causes of the distinct clinical presentations observed, although they still remain uncharacterized.^9,18 The authors additionally hypothesize that differences in the spinal cord vascular supply, speed of extrusion, and resistance of the spinal cord to contusion and compression may play a role in the clinical presentation of these animals, with further work needed to assess these factors.

The present study had a number of limitations, particularly those related to its retrospective nature and relatively small number of cases. The exclusion of breeds other than French Bulldogs assured high homogeneity of the study population but prevents extrapolation of the findings to other breeds. Furthermore, owing to patient recruitment methodology, our study population was biased toward younger patients (median age, 12 months) with more severe signs of brachycephalic obstructive airway syndrome, and it is possible that spinal cord-to-vertebral canal area ratios could change with age. The median age of French Bulldogs with intervertebral disk extrusion is 4 years.²¹ MRI would have been preferable in addition to CT for assessment of spinal cord area; however, CT has been previously used for the same purpose,^12,15 and MRI scans covering the cervical and thoracolumbar regions of the same dog are not commonly available because imaging is usually more centered on the region of interest.

Because CT was used, animals with subtle sites of spinal cord compression may have been included. This effect was limited by having 2 observers review the images individually and blinded. Additionally, a review of the patients’ medical records to include CT imaging reports prepared by a board-certified radiologist was made by one of the authors for further accuracy. The measurement technique itself is also inherently imprecise, and use of a bone window, as opposed to a lung window reconstructed to bone, would have increased the accuracy of polygon tracing and area measurement; however, this was not possible owing to insufficient patient imaging data. Even with this imprecision, the conclusion that the epidural space was greater in the cervical region was repeated by both reviewers. Furthermore, measurements were made at only 2 sites, which prevents extrapolation of the findings to the entire cervical and thoracolumbar regions, and at the level of C5, the cranial portion of the cervical intumescence may be included, thereby reducing the epidural space in that region.

In conclusion, contrary to previous reports, the findings of this study suggested that the epidural space is not larger in the cervical region than the thoracolumbar region, at least in this population of young French Bulldogs. Further research is indicated to confirm this finding, particularly in other breeds, and to elucidate the potential reasons for the often more severe neurologic deficits observed with thoracolumbar disk extrusions, compared with cervical disk extrusions.

Supplementary Materials

Supplementary materials are available at the journal website: avmajournals.avma.org