SEARCH

SEARCH BY CITATION

Keywords:

  • Allodynia;
  • Chemokines;
  • Dog;
  • Hyperpathia

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References

Background

Neuropathic pain can be a clinical sign in Cavalier King Charles Spaniels (CKCS) with syringomyelia. The pathophysiology of this pain is not fully understood.

Hypothesis

Neuropathic pain in CKCS is a result of a neuroinflammatory process.

Animals

Twenty-six client-owned dogs: 15 dogs with clinical signs of cervical hyperesthesia (group 1), and 11 dogs without of clinical signs (group 2).

Methods

Dogs were examined by magnetic resonance imaging (MRI). Interleukin-6, tumor necrosis factor alpha, and substance P were measured in CSF and compared with morphological findings on MRI and clinical pain scores.

Results

All dogs without clinical signs had symmetrical syringomyelia, whereas in the group with pain, 6 dogs had symmetrical and 9 dogs had asymmetrical syringomyelia. Pain and syringomyelia asymmetry were correlated, and a strong association between pain and dorsal horn involvement of syringomyelia was observed. There was no significant difference between the mean width of the syringomyelia in dogs with or without pain. The concentrations of interleukin-6 and substance P were significantly higher in dogs with neuropathic pain. Tumor necrosis factor alpha was not detected in either group.

Concentrations of substance P were significantly higher in dogs with asymmetrical syringomyelia or dorsal horn involvement, whereas interleukin-6 concentrations were not significantly different between groups.

Conclusion

Release of interleukin-6 and substance P may initiate proinflammatory effects leading to development of persistent pain in CKCSs with syringomyelia.

Abbreviations
AMPA

α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

CM1

Chiari malformation type 1

CLM

Chiari-like malformation

CKCS

Cavalier King Charles Spaniel

CSF

cerebrospinal fluid

FFE

fast field echo

FLAIR

fluid-attenuated inversion recovery

IU

international units

MRI

magnetic resonance imaging

NMDA

N-methyl-D-aspartate

A neurological condition comparable with Chiari malformation type 1 (CM1) in humans has been described in Cavalier King Charles Spaniels (CKCSs).[1] These morphological findings have been referred to as the Chiari-like malformation (CLM), which has been proposed to be the canine analog of the human disease.[2] Whereas CM1 in humans is characterized by a smaller caudal fossa volume with descent of the cerebellar tonsils through the foramen magnum, the identification of CLM in CKCS is based on morphological variables (eg, cerebellar herniation, indentation of the occipital bone). Therefore, it may be more correct to use an anatomical description (ie, caudal occipital malformation syndrome). This aberration from normal anatomy can be associated with fluid-filled cavities in the spinal cord parenchyma (syringomyelia).[2]

Syringomyelia can be associated with clinical signs of cranial and cervical hyperesthesia with characteristics of neuropathic pain.[3] Interestingly, magnetic resonance imaging (MRI) findings and clinical signs in affected dogs show limited correlation (ie, dogs with severe syringomyelia can have no clinical signs and dogs with minor syringomyelia can have severe clinical signs).[4, 5] The precise origin of the pain in clinically affected dogs is unknown. It has been found, however, that development of pain seems to be correlated with the width and asymmetry of the syrinx leading to destruction of the dorsal horn of the gray matter.[4] In contrast, syringomyelia in asymptomatic CKCSs has been found to be symmetrical on MRI and histopathological examination.[4, 5]

Injury to the spinal cord can cause neuropathic pain, which is not only a consequence of the morphological changes of the parenchyma but also of inflammatory changes.[6] Glial cells can be activated by cell damage and, similar to other immune cells, release proinflammatory cytokines such as interleukin-1, interleukin-6, and tumor necrosis factor.[7-12] Pro-inflammatory cytokines have been associated with exaggerated pain states, which may be mediated by altered expression of various neurotransmitters including substance P.[7-12] It has been proposed that cytokines and neuropeptides could be involved in neuropathic pain in CKCS with syringomyelia.[3] Thus, in the present study, we aimed to assess the possible role of interleukin-6, tumor necrosis factor alpha, and substance P in the pathophysiology of neuropathic pain in CKCSs by measuring the concentration of these mediators in the cerebrospinal fluid (CSF) and correlating them with morphological changes.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References

Between 2005 and 2012, 137 CKCSs were examined at the Clinic for Small Animals of the Justus-Liebig-University, Giessen, Germany. The medical records and MRI findings in these dogs were reviewed retrospectively. The medical history and results of general physical and neurological examination also were reviewed.

All CKCS underwent MRI of the head and cervical spine for breeding selection at the request of their owners. The dogs were sedated with 0.1 mL/kg diazepam1 and 0.3 mL/kg propofol2 given IV. Anesthesia was maintained after endotracheal intubation by inhalation of isoflurane3 in oxygen. MRI was performed on a 1.0 Tesla Scanner.4 T1- and T2-weighted sagittal and transverse images of the head were obtained. In addition, T2-weighted and fluid-attenuated inversion recovery (FLAIR) sagittal images of the spinal cord up to the level of the 5th thoracic vertebra were acquired. The detection of a hyperintens signal of ≥2 mm diameter within the spinal cord parenchyma in T2-weighted images, which was hypointense in T1 and FLAIR was consistent with the presence of syringomyelia. If syringomyelia was present, a transverse T2-weighted and a dorsal and transverse fast field echo (FFE) over the whole length of the spinal cord segment affected by syringomyelia was added. CSF was collected from all dogs with neurological disease as a standard operating procedure in our clinic unless owners explicitly declined. This procedure is in accordance with the universitary animal protection guidelines. CSF samples are divided and stored for potential additional diagnostic tests in plastic tubes,5 at −80°C. CSF puncture was performed at the cervicomedullary cistern. The medical history and presence of syringomyelia-associated pain were determined from the medical records. The images of 26 CKCS with syringomyelia were retrospectively evaluated. Selection of dogs was based on the results of CSF analysis and observation of the clinical signs (dogs must have been under surveillance for the entire day). Dogs were assigned to the pain group based on owners' descriptions of clinical signs of hyperesthesia. These included dogs that disliked touch to certain areas of the skin and that were reluctant to grooming or a neck collar. Scratching on one side only while walking (often without skin contact) and vocalizing during these episodes were defined as pain.[1] Absence of any other painful disease as indicated by physical and neurological examinations and the MRI findings was another inclusion criterion. Fifteen dogs showing clinical signs of cervical hyperesthesia and scratching were assigned to group 1, 11 dogs free of clinical signs were included in group 2. Images were exported to specialized visualization software6 for evaluation of syringomyelia. Syringomyelia was evaluated along its full longitudinal extent for asymmetry in transverse FFE images. Left/right asymmetry of syringomyelia was assessed using a vertical line through the middle of spinal cord and measuring the extension from the midpoint of that line into the largest expansion of the syringomyelia (Fig 1A). The maximal width of syringomyelia was measured on a transverse T2-weighted scan at the largest diameter of the syringomyelia (Fig 1B). Dorsal horn involvement was recorded based on subjective assessment of the topographical anatomy of the spinal cord in the FFE sequences (Fig 1C). All images were evaluated independently by two of the authors (MS, NO) and reevaluated if there were discordant results.

image

Figure 1. Transverse MR images showing syringomyelia morphology. Left/right asymmetry of the syringomyelia was assessed on a transverse T2 weighted MR image using a vertical line through the middle of the cord and measuring the extension from the midpoint of that line into the largest expansion of the syringomyelia (A). The maximal width of syringomyelia also was measured on a T2-weighted transverse scan (B). Dorsal horn involvement was recorded based on subjective assessment of the topographical anatomy of the spinal cord in the FFE sequences (C).

Download figure to PowerPoint

CSF Measurements of Interleukin-6, Tumor Necrosis Factor Alpha, and Substance P

Frozen CSF specimens were evaluated 4–34 months after sampling. Determination of interleukin-6 and tumor necrosis factor alpha was performed by bioassays based on dose-dependent growth stimulation of interleukin-6 on the B9 hybridoma cell line and on a cytotoxic effect of tumor necrosis factor alpha on the mouse fibrosarcoma cell line WEHI 164 subclone 13 as described previously.[13] The detection limit of the assay was 3 international units (IU)/mL for interleukin-6 and 6 pg/mL for tumor necrosis factor alpha.

A dog-specific ELISA Kit was used7 for analyses of substance P in CSF according to the instructions provided by the manufacturer. Samples were evaluated in duplicate, and the detection limit of the specific assay was 10 pg/mL. For statistical analyses, values for substance P concentrations below the detection limit (pain-free group, n = 11; pain group, n = 3) were defined as 0 pg/mL.

Statistical Analysis

All data were analyzed using a statistical software package.8 Normal distribution of the data was assessed by the Shapiro–Wilk test and additional appropriate tests were chosen on the basis of normal or non-normal distribution as appropriate.

Relationship between Syringomyelia Morphology and Pain

Chi-squared tests were used to test the association between pain and asymmetry of the syringomyelia as well as pain and extension of the syringomyelia into the dorsal gray matter. The difference between the relative syringomyelia width in the dogs with and without pain was evaluated using a Student's t-test.

Relationship between Mediators and Pain

A Student's t-test also was used to evaluate a possible difference between the mean amounts of substance P, interleukin-6, and tumor necrosis factor alpha in dogs with and without pain.

In addition, the correlation between interleukin-6 and substance P was tested using linear regression analysis.

Relationship between Morphology and Mediators

The putative association between the asymmetry of the syringomyelia, as well as the extension of the syringomyelia into the dorsal gray matter, and the amounts of tumor necrosis factor alpha, interleukin-6, and substance P were tested using Chi-squared tests. The correlation between the amounts of substance P, interleukin-6, tumor necrosis factor alpha and the maximum width of the syringomyelia was evaluated using linear regression analysis. P values <than .05 were considered statistically significant in all statistical tests (95% confidence interval).

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References

Epidemiologic Data

Of the 26 dogs examined, 10 dogs in group 1 were female and 5 male (age, 5 months to 9.5 years; mean, 3.9 years; weight range, 4.2–12 kg). In group 2, there were 6 female and 5 male dogs (age, 6 months to 8 years; mean, 4.0 years; weight range, 6–17 kg). Median pain score was 3 (range, 3–4). In the painful group, 11 dogs had a pain score of 3 (scratching and vocalization > once per week), and 4 dogs had a pain score of 4 (scratching and vocalization > once per week, activity compromised).[2]

MRI Examination

Both examiners agreed that all dogs without clinical signs had symmetrical syringomyelia (Fig 2A–C) whereas, in the group with pain, 5 dogs had symmetrical and 6 dogs had asymmetrical syringomyelia (Fig 2D–F). Left/right asymmetry was found only within the dorsal half of the spinal cord gray matter, into which all syringomyeliae lesions extended. There were no discordant results concerning the assessment of asymmetry.

image

Figure 2. The morphology of the syringomyelia in symptomatic (A–C) and asymptomatic dogs (D–F). In the dogs with clinical signs there is an association between pain and asymmetrical spinal cord cavitations extending in the gray matter of the spinal cord, whereas the syringomyelia is symmetrical in the dogs without clinical signs (white arrows).

Download figure to PowerPoint

Relationship between Syringomyelia Morphology and Pain

We found an association between pain and syringomyelia asymmetry (P = .0017; Phi = 0.64), and a strong association between pain and dorsal horn involvement of syringomyelia (P = .0002, Phi = 0.70). Dogs with pain had a mean syringomyelia width of 0.36 cm (SD ± 0.0276) compared to a mean of 0.34 cm (SD ± 0.02906) in dogs without pain. There was no difference in the mean width between the groups (P = .64).

Relationship between Mediators and Pain

The mean amount of interleukin-6 was 0 IU/mL in group 1 (no pain) and 50 IU/mL in group 2 (pain; Fig 2A). The mean concentration of substance P was 0 pg/mL in group 1 (no pain) and 12 pg/mL in group 2 (pain; Fig 2B). The concentrations of interleukin-6 and substance P were significantly higher in the pain group compared with the pain-free group (P < .0001 and P = .0002).

Relationship between Syringomyelia Morphology and Mediators

There was an association between dorsal horn involvement and the presence of interleukin-6 (P = .018), and between dorsal horn involvement and the presence of substance P in the CSF (P = .001). Asymmetry of the syringomyelia was not associated with interleukin-6 (P = .13) but with substance P (P = .001). Syringomyelia width was neither correlated with the amount of interleukin-6 (P = .79) nor with substance P (P = .839). The amount of substance P was correlated with the amount of interleukin-6 (= 0.491, P = .0075, Fig 3). Tumor necrosis factor alpha could not be detected in the CSF of the dogs.

image

Figure 3. Box and whisker plots illustrating the difference of the median interleukin-6 (A) and substance P (B) concentration in the cerebrospinal fluid of Cavalier King Charles Spaniels with and without clinical signs.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References

Dogs with clinical signs of pain (Fig 2) showed a certain syringomyelia morphology (large width, asymmetry, involvement of the dorsal horn gray matter), which has been found in a previous diagnostic imaging investigation.[4] In our study, only asymmetry and dorsal horn involvement were associated with pain. The expansion of syringomyelia can be an ongoing process over time.[14] The lack of significant differences in width of the syringomyelia between dogs with and without clinical signs could be influenced by the individual course of the development of structural changes in the parenchyma. If more dogs would have been examined at older ages, the differences might have been more pronounced. This could explain the lack of association between syringomyelia width and pain in our study.

image

Figure 4. Scatterplot illustrating the correlation between interleukin-6 and substance P in dogs with clinical signs of pain.

Download figure to PowerPoint

Furthermore, we found evidence that clinical signs in CKCSs with syringomyelia are not only associated with a certain syringomyelia morphology but also with the presence of interleukin-6 and substance P in the CSF of clinically affected CKCS.

Substance P belongs to a group of neurokinins that are broadly distributed in the central nervous system with evidence for both neuronal and glial cells as being sources of this neuropeptide.[15] There is evidence for the role of substance P in regulating the immune functions of spinal glial cells. Activation of substance P (neurokinin-1) receptors elicits signal transduction pathways in both cell types that can initiate or augment inflammatory responses in astrocytes and microglia. Release of substance P is induced by trauma and other stimuli to the spinal neurons. Interleukin-6 is a key component of the nervous system's response to injury.[16, 17] After neuronal destruction, interleukin-6 normally initiates a cytotrophic reaction. Therefore, the physiological benefit of interleukin-6 release would be activation of the tissue response to cellular damage initiating regeneration.

Despite their positive effects in neuronal protection and regeneration, cytokines also can be involved in the development of neuropathic pain by direct modulation of dorsal horn neuron activity. Both interleukin-6 and substance P can not only acutely excite primary sensory neurons but also lead to a sustained increase in their excitability.[18] Interleukin-6 can increase the conductivity of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) receptors, and also can upregulate the number of these receptors on the surface of neurons.[19] This leads to increased electrical afferent input into the dorsal horn that enhances the “central sensitization” process.[20, 21]

Interleukin-6 also has an indirect effect on the hyperexcitability of adjacent pain transmitting neurons by enhancing the release of substance P and other neuropeptides from presynaptic primary afferent terminals.[10] Among the pathological effects of substance P is the ability to potentiate the excitation of nociceptive neurons leading to prolonged postsynaptic excitations.[22, 23] Neuronally derived substance P, in turn, can contribute to the release of chemokines from glial cells which can lead to exacerbation of the inflammatory process. Understanding of neuropathic pain in the CKCS formerly was focused primarily on neuronal mechanisms. It was proposed to be caused by damage to the spinothalamic tract,[3, 6, 24, 25] deafferentation and denervation with hyperexcitability of neurons,[6] disinhibition or imbalance between pathways, as well as reorganization, of primary afferents and remapping of central neurons.[26-29] However, the presence of substance P and interleukin-6 in the CSF of CKCSs with pain suggests a role of activated glial cells as modulators of pain after cell damage in these dogs. Activated glial cells also have been identified in histopathological examination of the spinal cord of CKCSs with syringomyelia.[24]

We could not detect tumor necrosis factor alpha in either group of dogs. Proinflammatory cytokines often are sequentially released in a cascade in which tumor necrosis factor alpha typically is produced first, causing the induction of interleukin-1, which in turn causes the induction of interleukin-6.[30] Thus, tumor necrosis factor alpha is usually detected in acute inflammatory conditions, which most likely explains the absence of tumor necrosis factor alpha in the CSF specimens of dogs with syringomyelia, known to be a chronic process. Our results also confirm previous findings of others, which also only detected increased interleukin-6 but no tumor necrosis factor alpha in CSF of different pain-related conditions, including regional pain syndrome or lumbar radiculopathy.[31-33]

The relationship between syringomyelia morphology and increased concentrations of interleukin-6 and substance P remains elusive. Asymmetrical syringomyelia also has been found to produce clinical signs in human patients,[34] and patients with asymmetrical syringomyelia seem to be less responsive to surgical intervention.[35] Symmetrical syringomyelia on the other hand also does not produce clinical signs in human patients,[35] which supports the role of asymmetry in the pathogenesis of pain in both humans and dogs. Sections from the spinal cord of symptomatic CKCSs have reinforced the role of asymmetrical syringomyelia deviating into the dorsal horn in these dogs.[24] CKCSs with pain show greater degrees of syrinx-associated tissue loss and degeneration of neurons in the gray matter accompanied by proliferation of astrocytes. Syringomyelia margins were well defined and sometimes lined with collagen in symptomatic dogs. In contrast, syringomyelia in asymptomatic CKCSs has not only been found to be symmetrical, but the parenchyma adjacent to the syringomyelia was spongiotic and edematous. Reactive glial cells were not found in the dogs without clinical signs of pain, and adaptive reactions of the parenchyma were lacking.[24] This difference in parenchymal reaction between dogs is striking. It remains unclear why some dogs respond with repair and adaption to the mechanical forces of the expanding syringomyelia and others do not. Different mechanisms in the pathogenesis could perhaps determine outcome and the clinical features of the condition.[35] The extent of pressure in the central canal, the onset and duration of increased pressure and its fluctuations might be influential factors for different tissue reactions and distinct cavitary patterns.

Conclusion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References

The release of interleukin-6 and substance P may initiate proinflammatory responses with pathological effects leading to neuronal hyperexcitability. We suggest that release of these substances is 1 factor in development of persistent pain in CKCSs with syringomyelia. The recognition of the involvement of inflammatory mediators in this process could offer new treatment options for CKCS, and detection of CSF interleukin-6 and substance P concentrations might be a new diagnostic tool to identify potential treatment strategies.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References

We thank J. Hermann for advice on statistical analyses and J. Murgott for technical assistance.

Conflict of Interest Statement: None of the authors have a financial or personal relationship with other people or organizations that could inappropriately influence or bias the content of the paper.

Footnotes
  1. 1

    Diazepam, rathiopharm GmbH, Ulm, Germany

  2. 2

    Propoflo, Ahlbrecht, Aulendorf, Germany

  3. 3

    Isoflo, Ahlbrecht

  4. 4

    Phillips Brilliance, Phillips, Amsterdam, Netherlands

  5. 5

    Eppendorf tubes, Eppendorf Ag, Hamburg, Germany

  6. 6

    Synedra IT GmbH, Berlin, Germany

  7. 7

    E90393Ca, USCN Life Science Inc, Houston, TX

  8. 8

    Graph Pad Prism 4.0, Graph Pad Software Inc, San Diego, CA

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgments
  8. References