Spasticity is a commonly recognized complication following SCI in humans, affecting an estimated 60% to 78% of people living with chronic SCI.1,2 Only rare reports3–5 exist involving spasticity due to experimentally induced SCI in dogs, and no data are available regarding the frequency and severity of spasticity in dogs with naturally occurring SCI.
In human medicine, the most frequently cited definition refers to spasticity as a motor disorder characterized as a velocity-dependent increase in tonic stretch reflexes (muscle tone) with exaggerated tendon jerks due to hyperexcitability of the stretch reflex.6 Clinical spasticity can most accurately be described as a combination of abnormalities, including exaggerated reflexes, increased muscle tone, uncontrolled limb movements, and flexor and extensor spasms, any of which may be more or less apparent in an individual patient and may develop secondary to > 1 pathophysiologic mechanism.7,8 Proposed mechanisms include a combination of factors related to loss of supraspinal input and maladaptive central processing of afferent sensory input to the spinal cord below the level of the SCI, which together cause an increase in motor neuron excitability and the clinical manifestations of spasticity.8
Although clinical findings of spasticity in affected limbs are simple to recognize, the condition represents a complex problem and accurate quantification is challenging. Several clinical scales have been developed in human medicine to quantify the severity of spasticity, including the Ashworth Scale (also the Modified and Modified Modified Ashworth Scale) and SCATS.9–14 The Ashworth Scale is used to assess an increase in muscle tone and resistance to passive movement and provides a global measure of spasticity, whereas the SCATS is used to evaluate clonus, flexor spasms, and extensor spasms.9,10 Additional evaluation methods include self-reported measures such as the Penn Spasm Frequency Scale as well as biomechanical and electrophysiologic testing.2,15–23
The clinical scales are straightforward to implement in a clinical setting, but no single test provides a comprehensive evaluation and so multiple tests are often combined. Clinical tests for spasticity are also variably correlated with each other, with patient-reported impairment attributable to spasticity, and with electrophysiologic and biomechanical measurements.2,10,11,15,19,21,23,24 Furthermore, spasticity varies between and within affected individuals over time.14,15,21 Nevertheless, spasticity can cause pain, interfere with daily functioning, and adversely affect quality of life for humans with SCI.1,15,21,25–28 The impact of spasticity coupled with the high prevalence in this population warrants its use as the primary outcome measure in clinical trials and development of improved diagnostic testing methods.29
Dogs with severe, chronic thoracolumbar SCI commonly develop signs of spasticity weeks to months after the initial injury.3,5,30 However, no clinical assessment tools exist in veterinary medicine for the evaluation of spasticity in affected dogs. The impact of spasticity on affected dogs is also unknown. The purpose of the study reported here was to use the aforementioned human scales to develop a feasible and repeatable scale for assessment of spasticity in chronically paralyzed dogs and to investigate factors associated with spasticity development and the relationship between spasticity and gait.
Materials and Methods
Ethics statement
Informed consent was obtained from all owners of participating dogs. The study protocol was approved by the North Caroline State University Institutional Animal Care and Use Committee (protocol No. 15-004-01).
Dogs
Dogs were prospectively enrolled from among patients of the Canine Spinal Cord Injury Program at the North Carolina State University College of Veterinary Medicine between March and December 2015. For inclusion, dogs were required to have chronic gait deficits with absent or reduced hind limb and tail pain perception (with or without urinary and fecal incontinence) due to an acute thoracolumbar (T3-L3 region) SCI sustained a minimum of 3 months previously. Data were collected for each dog regarding signalment, concomitant medications, diagnosis, lesion location, injury duration, and SCI treatments.
General neurologic evaluation
All dogs underwent an initial neurologic examination that included evaluation of gait, proprioception, spinal reflexes, and pain perception. Gait analysis consisted of watching the dog walk on a nonslip surface as well as on a treadmill, with sling support provided if it was unable to walk unassisted. The treadmill was adjusted to a comfortable speed for each dog, and dogs were walked for approximately 3 minutes. All examinations were videotaped. Gait was categorized as ambulatory (at least 10 consecutive weight-bearing steps unassisted), nonambulatory with motor function, or paraplegic and was also categorized with previously validated objective gait assessment tools (OFS, SS, and RI; Supplementary Appendix S1, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.78.7.854).31–34
For proprioceptive limb placement and hopping tests, findings were recorded as absent (0), delayed (1), or normal (2). For patellar and withdrawal reflex tests, findings were recorded as absent (0), decreased (1), normal (2), increased (3), or clonus (4). For evaluation of hind limb muscle tone, findings were recorded as decreased or flaccid, normal, or increased. For the cutaneous trunci reflex, the vertebral level of the caudal border where the reflex could still be elicited was recorded. Nociception in the hind limb and tail was recorded as present or absent, and urination was recorded as voluntary or involuntary.
Initial spasticity testing
All dogs received a spasticity evaluation at the time of neurologic examination. Tests of muscle tone, clonus, and flexor and extensor spasms were adapted from human clinical scales (Ashworth Scale and SCATS) and evaluated for feasibility.10,11 The Ashworth Scale involves use of resistance to passive movement to assess increases in muscle tone, and responses are scored on an ordinal scale for each muscle group evaluated. The SCATS involves evaluation of plantar flexor clonus, hind limb flexor spasms, and hind limb extensor spasms, with each quantified on an ordinal scale ranging from 0 to 3. Clonus of the plantar flexors is scored on the basis of the duration of clonic activity following rapid, passive dorsiflexion of the ankle joint. Flexor spasms are graded by duration and degree of response following application of pinprick stimulation to the bottom of the foot. Extensor spasms are elicited by passive, simultaneous extension of the stifle and hip joints and scored by duration of the resultant, visible quadriceps contraction.
For feasibility testing, each dog was positioned in lateral recumbency and lightly restrained to ensure it remained still and relaxed, and the uppermost limb was tested. First, resistance to passive range of motion was assessed at each of the hock, stifle, and hip joints (adapted Ashworth Scale). For each joint, observed tone was categorized as normal (no resistance) or mildly, moderately, or markedly increased (passive movement difficult). Following testing of the first hind limb, the dog was rotated in position, and the other hind limb was tested similarly. Results obtained with the adapted Ashworth Scale were extremely variable in all dogs, so this scale was no longer considered for inclusion in the CSS.
While dogs were still positioned in lateral recumbency, patellar and plantar flexor clonus and flexor and extensor spasms were then assessed (adapted SCATS). The patellar reflex was elicited by striking the patellar tendon with a reflex hammer to assess for the presence of clonus. With the limb relaxed in extension, flexor spasms were triggered by pinprick stimulation of the skin of the bottom of the foot with a 25-gauge needle at the midmetatarsal level. The stifle and hip joints were then flexed to 90°, and extensor spasms were assessed by monitoring for quadriceps contraction following simultaneous extension of the hip and stifle joints by the investigator. Extensor spasms were rarely elicited, so this particular test was eliminated from consideration for inclusion in the CSS. Plantar flexor clonus, tested by rapid dorsiflexion of the hind foot, was never elicited and this test was eliminated from consideration as well.
Final CSS and scoring
Assessment of patellar clonus duration and flexor spasm duration and degree was feasible in both hind limbs of all dogs, and these variables were used to create the CSS (Supplementary Appendix S2, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.78.7.854). Each dog was evaluated for patellar clonus and flexor spasm by use of the CSS scoring system. Tests were repeated 3 times (ie, 3 trials) on each hind limb, with at least 30 seconds separating trials.
Duration (as a continuous variable, in seconds) was recorded for each trial of clonus and spasm. If clonus or spasm was still evident 60 seconds after the trial was initiated, that trial was discontinued and a maximum duration of 60 seconds was recorded. The mean duration of the response (mean of the 3 trials) was then calculated for patellar clonus and flexor spasms in each limb. Individual durations for each trial and mean duration for each hind limb were subsequently converted to an ordinal scale, with a score of 0 indicating no clonus or spasm, a score of 1 indicating a duration of ≤ 3 seconds (mild), a score of 2 indicating > 3 but ≤ 10 seconds (moderate), and a score of 3 indicating > 10 seconds (severe).
Continuous and ordinal data for flexor spasm duration averaged across the 3 trials were compared in a subset of dogs for which the data were available (n = 11). These data were qualitatively similar. However, the distribution of continuous data was not normal, with most durations < 10 seconds and a broad range in duration among spasms lasting > 10 seconds, up to the 60-second cutoff. Given the skewed distribution but natural separation into categories, ordinal data classification was used for the CSS data in the same manner as for the SCATS data.
Angle of flexion of the hip and stifle joints during flexor spasms was estimated and broadly categorized, with the 2 joints considered as 1 entity. By this system, a score of 0 indicated no flexion, a score of 1 represented < 10° (mild), a score of 2 represented 10° to 45° (moderate), and a score of 3 represented > 45° (severe). The median score for the 3 trials was used to assign the overall score for degree of flexor spasms for each hind limb.
The mean (clonus and spasm duration) or median (spasm degree) score (0 to 3) was assigned for each hind limb for each scale component, which were then summed for the right and left hind limbs to yield a total score (0 to 6) for each component and an overall score representing all components and both hind limbs (0 to 18). Overall scores were also used to categorize dogs as having absent to mild spasticity (0 to 6) or moderate to severe spasticity (7 to 18).
To evaluate the variability in spasticity (CSS scores) over time within a given dog, a subset of dogs (n = 10) was returned for repeated testing by the same rater. The same testing protocol (for both hind limbs) was applied at each testing session. Testing was repeated once for 8 dogs and twice for 2 dogs, with the period between the initial and subsequent testing sessions ranging from 48 hours to 7 months. Interrater reliability was determined by having 2 raters (one of whom had no involvement in scale development) test a subset of dogs on the same day and in the same testing conditions, with 2 to 5 minutes separating testing sessions.
Owner questionnaire development
A questionnaire was developed to collect from dog owners information regarding the presence, frequency, and impact of spastic movements in their dogs at home in a manner comparable to the Penn Spasm Frequency Scale and measures of quality of life for people living with SCI (adapted from the Penn scale; Supplementary Appendix S3, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.78.7.854).19,20,24,27,28,35 Requested information included presence (yes or no) of involuntary movements, jerking, or muscle spasms (in flexion, extension, or both); frequency of these movements (> 10 times/d, 1 to 10 times/d, > 1 time/wk, or < 1 times/wk); time of day at which these movements were noticed (day, night, or random); asymmetry of these movements (left hind limb, right hind limb, or random); and impact of these movements on daily functioning or quality of life (none, minimal, moderate, or severe). Because of the subjectivity inherent to an owner-reported measure, spastic movements at home were classified as present or not present for comparison with observations recorded for the investigator-administered CSS.
Statistical analysis
Statistical softwarea,b was used for all analyses. The CSS overall score and its 3 individual component mean or median scores were summarized for all dogs and are reported as median (range). Interrater reliability and temporal variation in measurements for dogs that were evaluated more than once were measured by calculation of intraclass correlations. To identify factors that may influence the development of spasticity, associations were examined between overall CSS score and dog age, SCI duration, and lesion location (cranial or caudal to T13, to capture the possible effect of greater numbers of intact spinal segments on the development of spasticity) by use of logistic regression (age and injury duration) and 1-way ANOVA (lesion location). Logistic regression and the Wilcoxon rank sum test were also performed to determine whether CSS scores (overall and individual component scores) were associated with the ability to walk (yes or no) and gait scores (derived from the OFS, SS, and RI). The Holm-Bonferroni method was used to correct P values for multiple comparisons (denoted as Padjusted). For all analyses, values of P < 0.05 were considered significant.
Results
Dogs
Twenty dogs with thoracolumbar SCI were included in the study (Supplementary Appendix S4, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.78.7.854). Mean (SD) body weight was 13.1 kg (9.9 kg), and mean age was 6.0 years (2.7 years). Breeds included Dachshund (n = 6 [30%]), mix (3 [15%]), Dachshund-Chihuahua cross (2 [10%]), pit bull–type (2 [10%]), Miniature Schnauzer (1 [5%]), Labrador–Retriever-Poodle mix (1 [5%]), German Shepherd Dog (1 [5%]), English Bulldog (1 [5%]), Miniature Poodle (1 [5%]), Shih Tzu (1 [5%]), and Boston Terrier (1 [5%]).
Median SCI duration was 12 months (range, 4 to 84 months). Suspected or confirmed intervertebral disk disease was the most common diagnosis (n = 14). In all dogs, the site of the neurologic lesion was identified between T3 and L3, except for 1 dog with concurrent, less pronounced caudal cervical abnormalities. Magnetic resonance imaging or CT was performed for 15 (75%) dogs, revealing a spinal cord lesion at T13 or cranially in 8 dogs, caudal to T13 in 6 dogs, and diffusely distributed within the thoracolumbar portion of the spinal cord in 1 dog with a suspected inflammatory reaction secondary to an IM (epaxial) melarsomine injection.
At the time of evaluation, 7 (35%) dogs were paraplegic, 8 (40%) were nonambulatory with motor function, and 5 (25%) were ambulatory. Median OFS was 1 (range, 0 to 9), median unsupported SS was 0 (range, 0 to 89), and median unsupported RI was 0 (range, 0 to 46.56). Eighteen (90%) dogs had no signs of hind limb and tail pain perception, whereas 2 (10%) dogs had a severely blunted response in some digits of the hind limbs. Seventeen (85%) dogs were urinary incontinent, whereas 3 (15%) had consistent urine voiding without the need for manual bladder expression. No dogs had signs of pain on palpation of the vertebral column. No dogs were receiving anti-spasticity treatment when evaluated.
CSS
The CSS assessment was easy to perform and well tolerated by all dogs. Hind limb flexor spasms were extremely variable among dogs and were sometimes dramatic in their intensity and duration (Supplementary Videos S5 and S6, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.78.7.854). Patellar clonus was uncommon (identified in ≥ 1 hind limb in 5 [25%] dogs) and relatively brief (median score, 0; range, 0 to 4), compared with the frequency of flexor spasms (identified in all 20 [100%) dogs) and the duration of flexor spasms (median score, 2; range, 2 to 6). Median flexor spasm degree score was 4 (range, 1 to 6). Median overall CSS score was 7 (range, 3 to 11). Ten (50%) dogs were categorized as having absent to mild spasticity (overall score, 0 to 6), and the other 10 (50%) were categorized as having moderate to severe spasticity (overall score, 7 to 18; Table 1).
Gait scores and clinical characteristics of 20 dogs with chronic thoracolumbar SCI categorized by spasticity severity.
| CSS spasticity severity category | Median (range) CSS score | Mean (SD) age (y) | Mean (SD) SCI duration (mo) | No. of ambulatory dogs | Median (range) OFS | Median (range) SS | Median (range) RI |
|---|---|---|---|---|---|---|---|
| Absent to mild (n = 10) | 4 (3–6) | 5.2 (3.0) | 17.8 (19.5) | 1 | 0.5 (0–6) | 0 (0–36) | 0 (0–5.88) |
| Moderate to severe (n = 10) | 10 (8–11) | 6.7 (2.3) | 22.4 (24.7) | 4 | 4.0 (0–9) | 48 (0–89) | 9.62 (0–89) |
Possible range of overall CSS scores was 0 to 18, with scores of 0 to 6 representing absent to mild spasticity and 7 to 18 representing moderate to severe spasticity.
Flexor spasm duration and degree scores were significantly (P < 0.001) moderately to highly correlated with overall CSS score (R2 = 0.57 and R2 = 0.82, respectively) and were mildly correlated with each other (R2 = 0.33 [P = 0.008]). Patellar clonus duration scores were poorly correlated with flexor spasm duration and degree scores (R2 = 0.01 [P = 0.67] and R2 = 0.14 [P = 0.11], respectively) and with overall CSS score (R2 = 0.24 [P = 0.03]).
Eight dogs had a 1-point difference and 2 dogs had a 2-point difference between total score (all 3 components) for the left hind limb and total score for the right hind limb. Of these 10 dogs, 8 had no detectable asymmetry on clinical neurologic examination, whereas in 2 dogs, the limb with the more severe spasticity (characterized by greater duration or degree of flexor spasms) was the same as the more severely affected side on examination. Subtle to mild clinical asymmetry was noted during neurologic examination of an additional 6 dogs; however, in none of these dogs did CSS scores differ between left and right limbs.
CSS reliability and temporal variation
Interrater reliability for 2 raters both scoring 8 dogs was high (intraclass correlation coefficient, 0.93). Repeated testing of 10 dogs by the same rater identified some variability between testing times (intraclass correlation coefficient, 0.72). With total CSS score for each limb used, a difference between testing times of 1 point was observed for the left hind limb of 7 dogs and the right hind limb of 6 dogs. For only 1 dog did total scores differ by > 1 point, with a difference of 2 points in each hind limb observed between testing sessions held 3 days apart. Variability between testing times in flexor spasm duration was comparable to that for overall CSS scores.
Associations of CSS scores with other variables
No significant associations were identified between overall CSS scores and dog age (Padjusted = 0.29), body weight (Padjusted = 0.28), SCI duration (Padjusted = 0.93), and lesion location (cranial vs caudal to T13; Padjusted = 1.00; n = 20). Data regarding 2 dogs with atypical lesions (1 dog with concurrent injury in the C6-T2 region and 1 dog with diffuse thoracolumbar involvement) were removed from analyses involving comparisons of CSS scores with gait scores because their injuries may have impacted motor function differently than in dogs with focal lesions in the T3-L3 region. For the remaining 18 dogs, no significant (Padjusted = 0.26) association was identified between overall CSS score and ability to walk (yes or no). However, overall CSS score was significantly positively associated with OFS (Padjusted = 0.048), SS (Padjusted = 0.042), and RI (Padjusted = 0.048). To determine which aspect of spasticity was driving this relationship, the scores for each scale component (combined for both hind limbs) were compared with gait scores. No significant (Padjusted ≥ 0.38) associations were identified between patellar clonus duration scores or flexor spasm degree scores and any gait scores. However, flexor spasm duration scores were positively associated with the ability to walk (Padjusted = 0.01), OFS (Padjusted = 0.008), SS (Padjusted < 0.001), and RI (Padjusted < 0.001; Supplementary Video S7, available at http://avmajournals.avma.org/doi/suppl/10.2460/ajvr.78.7.854).
Owner questionnaire
Involuntary hind limb jerking movements at home were reported by owners for 16 (80%) dogs. For 10 of these dogs, spastic jerking reportedly occurred 1 to 10 times/d, whereas for 5 dogs, it occurred ≤ 1 time/wk, and for 1 dog, it occurred > 10 times/d. Four dogs were reported to have spasms, which reportedly occurred 1 to 10 times/d for 3 dogs and ≤ 1 time/wk for 1 dog. When present, spastic movements of any kind were considered asymmetric between limbs for 5 dogs, and 4 owners reported that spasticity varied with time of day (day vs night).
Owners of 18 (90%) dogs reported that additional spastic movements could be induced by tactile stimulation, such as during urinary bladder expression or touching the digits or hind limbs. Two owners reported a mild, adverse impact of spasticity on their dog's quality of life and daily functioning, whereas no impact was reported for the remaining dogs. No relationships were identified between spasticity observed at home and overall CSS scores categorized by severity. Of the 12 dogs for which owners reported spastic movements (involuntary jerking or spasms) occurred ≥ 1 time/d, 7 were scored as having absent to mild spasticity (overall CSS score, 0 to 6) and 5 were scored as having moderate to severe spasticity (overall CSS score, 7 to 18). Among the 8 dogs in which spasticity at home was reported less frequently or absent, 3 had absent to mild spasticity as measured with the CSS and 5 had moderate to severe spasticity.
Discussion
Results of the present study suggested that spasticity in dogs with chronic, severe SCI can be quantified through assessment of patellar clonus and flexor spasms. Flexor spasms were the most prominent finding and varied widely among dogs, with spasm duration strongly associated with gait scores. The semiquantitative scale we developed may be a useful adjunct to the standard neurologic examination for dogs with incomplete recovery from thoracolumbar SCI as well as an important tool for investigating this largely ignored phenomenon in dogs in more depth.
Spinal cord injury is common in dogs, and the veterinary literature commonly refers to spasticity in dogs concurrent with other signs of upper motor neuron dysfunction.36,37 However, no consensus exists on the precise definition of spasticity in dogs and no validated, objective evaluation tools are available in veterinary medicine. We developed a noninvasive test from clinical scales currently used in human medicine,9–14 and we focused on a sample of dogs that had sustained acute, functionally complete thoracolumbar SCI because disruption of supraspinal input is considered an important component of spasticity.8 To capture a range of spasticity, the included dogs had a wide range in hind limb motor function despite failure to recover normal hind limb and tail pain perception. We elected to include 2 dogs with blunted but not completely absent pain perception, 1 dog with a diffuse thoracolumbar injury, and 1 dog with multifocal injury because these dogs had profound neurologic deficits indicative of severe injury and an incomplete recovery. They were, therefore, at risk for spasticity and were part of the clinical population we wished to investigate. The dog with diffuse thoracolumbar injury and the dog with multifocal injury were removed from statistical analysis of relationships between CSS scores and gait scores because the impact of their lesions on motor function (compared to the more typical, focal lesion) was unknown.
During development of the spasticity testing protocol in the present study, extensor spasms were elicited rarely from the dogs, which is distinct from similar testing in humans, in whom extensor spasms are more common than plantar clonus.10 Patellar clonus was also uncommon in the dogs of the present study but was still consistently identified in 25%. In contrast, flexor spasms were elicited to at least some degree in all dogs and were severe and prolonged in some dogs. The reason for these discrepancies from findings in humans is unclear but may reflect positioning differences. Flexor and extensor spasms are assessed in humans while lying supine, whereas many dogs do not appear comfortable or are not stable lying on their backs.10 Instead, we positioned dogs in lateral recumbency, allowing them to lie comfortably and remain still throughout the testing period, with their hind limbs easily accessible and relaxed in relative extension.
Neuroanatomic differences between species may also explain the differences in frequency of clonus and flexor and extensor spasms between the study dogs and humans. For example, the corticospinal tract is the primary descending motor pathway controlling voluntary movement in humans but is relatively less important in dogs, compared with the rubrospinal and reticulospinal tracts.36,38,39 Although overlap in function and compensation after injury between these upper motor neuron tracts has been demonstrated in multiple species, the relative importance of disruption of the corticospinal tract in humans versus dogs, rather than just the general loss of supraspinal input on α motor neurons, may result in variations with regard to when, how, and to what extent spasticity develops in these 2 species.39–41 Despite flexor spasms being the most prominent component and patellar clonus being infrequently elicited and poorly correlated with the other components, we elected to maintain all aspects as a combined scale for the purposes of reporting our initial CSS for dogs with SCI. In this patient population, flexor spasm duration alone might prove most useful and practical in the evaluation of spasticity. However, patients with different SCIs may have different components of spasticity, including extensor spasms.
The final spasticity testing protocol required minimal training to perform, and results were comparable when testing was performed by 2 raters with different clinical backgrounds. Because spasticity within an individual is not a static feature, we assessed the temporal variability in test results for 10 dogs. We noted a mild to moderate difference in scores for most dogs evaluated more than once, consistent with fluctuations of spasticity testing results reported for humans.21 However, re-testing was performed only once for 8 of 10 dogs, so this preliminary finding may not represent the true temporal fluctuation in spasticity in dogs. Repetitive testing on multiple occasions may provide further information about the extent of variation in spasticity in dogs with chronic SCI, the degree to which the CSS captures this vacillation, and how results compare with those for humans with similar injuries.
The owner questionnaire was developed in the present study to capture observations on the manifestation of spasticity in dogs in their home setting. Findings obtained with the veterinarian- or investigator-administered CSS were not associated with owner reports of spasticity in their dogs at home, thereby mirroring the variable correspondence observed between physician- or investigator-administered scales (such as the Ashworth Scale) and patient-reported scales (such as the Penn Spasm Frequency Scale).19,24 These findings, considered with the temporal variation in CSS scores, might highlight the importance of assessing dogs with chronic thoracolumbar SCI more than once and might reflect the dynamic nature of spasticity.
In the owner questionnaire, responses were requested in the form of frequency categories, but in working with the owners, we found that spastic movements were not something they were consciously tracking prior to study participation, thereby limiting the reliability and applicability of these preliminary results. Additionally, the dog owners rarely perceived spasticity to have an adverse impact on their dog's daily functioning or quality of life. This may, in fact, indicate that the spasticity was not problematic secondary to the chronic SCI, that owners failed to appreciate the adverse impacts of spasticity, or that the questions failed to elicit information regarding these adverse impacts. Owners whose dogs had the weakest hind limb motor function generally appeared more aware of spastic limb movements than did owners of dogs that could walk (but might still have had spasticity). Further development and validation of the questionnaire, including longitudinal collection of owner observations and CSS measurements from the time of SCI, are warranted to investigate the contribution of such information to the clinical assessment of spasticity in dogs with SCI.
An ordinal scoring scale adapted from human scoring systems was chosen for use in the present study. It is possible that, for patellar clonus and flexor spasm durations, use of continuous data may more accurately depict the degree of spasticity at the point of testing in dogs with chronic thoracolumbar SCI, although increasing the accuracy of angle and duration measurement would need to be addressed, perhaps by careful analysis of video-recorded testing sessions. However, data in continuous and ordinal format were largely similar for the study dogs. Given the fairly small number of dogs and the skewed nature of the continuous data (toward dogs with the most severe spasticity), allocation to limited categories facilitated statistical comparisons between groups and variables. This use of broad scoring categories would also facilitate ease of implementation in the clinical setting.
We focused on the overall CSS score but also examined each of the 3 scale components individually because we believed each component likely reflected different components of what is collectively referred to as spasticity. Flexor spasms appeared to be the most prevalent and useful component in the study dogs. Indeed, in different patient populations, other components of the CSS may be relevant and the scale could be adapted as needed. The positive relationship between flexor spasm duration and ability to step in the dogs with chronic thoracolumbar SCI suggested that development of flexor spasms might reflect an increase in the excitability of the intraspinal circuitry and could be linked to the recovery of stepping in such dogs. Because we focused on dogs with chronic injury and stable neurologic status, the CSS is not intended to be used to predict motor recovery. However, the effect of flexor spasms on functional ambulation warrants further investigation because this might be contrary to the anticipated finding that extensor spasms would facilitate walking. As noted previously, various possible explanations exist for the lack of extensor spasms in the study dogs, but it has also been demonstrated in animals with experimentally induced injury that recovery of standing (which predominantly involves antigravity extensor muscles) can have an adverse impact on stepping.42 The relationship between walking and spasms is likely more complex with the development of both flexor and extensor spasticity as well as recovery of stepping, all reflecting overlapping postinjury spinal cord changes.
Given the subjectivity inherent to use of a clinical scale, further validation of the CSS and testing protocol is needed. We demonstrated reliability between 2 raters assessing a small number of dogs, but additional testing of a larger number of dogs, by more raters, and in conjunction with more objective assessment methods such as electrodiagnostic techniques is warranted to determine whether the CSS would be useful for a broader variety of SCI patients and clinical settings. This testing might also provide additional information regarding the relationship, if any, between spasticity and dog and injury characteristics such as age, body weight, injury duration, or lesion location. We specifically evaluated spasticity in a group of chronically injured dogs with static neurologic status. Use of the CSS during the acute phase of injury and repeating it over time, however, might provide additional, indirect information regarding the intraspinal circuitry changes that develop following SCI and that underlie the development of spasticity. Such information may offer insight into the plasticity of the injured spinal cord as dogs transition from the acute to chronic injury phase and may be useful in designing more effective and consistent test components.
Despite any limitations, we believe that the CSS can be readily integrated into routine neurologic evaluation. Dogs with chronic SCI are frequently considered a uniform population, even though potentially clinically relevant differences exist. For example, all dogs in the present study had a similar severity of initial injury (functionally complete SCI) and none regained normal pain perception but had a highly variable degree of motor recovery (paraplegic to ambulatory) and CSS findings (which varied from mild to severe). Using the CSS, we were able to demonstrate that spasticity and, specifically, flexor spasms were significantly associated with validated measures of gait, including OFS and hind limb stepping and coordination. This suggests that the CSS is a tool worthy of continued development and may be useful to differentiate among dogs with permanent impairment following SCI and guide a personalized medicine approach to treatment strategies with potential application to humans.
Overall, we found that the CSS was a simple tool for quantification of spasticity in dogs with chronic thoracolumbar SCI, and severity of flexor spasms and overall CSS score were associated with gait scores. We conclude that the CSS may provide pertinent clinical information when added to the standard physical and neurologic evaluation of similar patients. Furthermore, the CSS may be a useful outcome measure in clinical trials involving dogs with chronic thoracolumbar SCI, although the impact of spasticity in dogs remains to be determined.
Acknowledgments
Supported in part by the T32 Comparative Medicine and Translational Research Training Program of the National Institutes of Health and North Carolina State University and the North Carolina State University Research and Innovation Seed Funding Program.
ABBREVIATIONS
| CSS | Canine spasticity scale |
| OFS | Open field score |
| RI | Regularity index |
| SCATS | Spinal cord assessment tool for spastic reflexes |
| SCI | Spinal cord injury |
| SS | Stepping score |
Footnotes
SAS, version 9.4, SAS Institute Inc, Cary, NC.
Jmp 12 Pro, SAS Institute Inc, Cary, NC.
References
1. Maynard FM, Karunas RS, Waring WP III. Epidemiology of spasticity following traumatic spinal cord injury. Arch Phys Med Rehabil 1990; 71: 566–569.
2. Sköld C, Levi R, Seiger A. Spasticity after traumatic spinal cord injury: nature, severity and location. Arch Phys Med Rehabil 1999; 80: 1548–1557.
3. Blauch B. Spinal reflex walking in the dog. Vet Med Small Anim Clin 1977; 72: 169–173.
4. Hall PV, Smith JE, Campbell RL, et al. Neurochemical correlates of spasticity. Life Sci 1976; 18: 1467–1471.
5. McBride WJ, Hall PV, Chernet E, et al. Alterations of amino acid transmitter systems in spinal cords of chronic paraplegic dogs. J Neurochem 1984; 42: 1625–1631.
6. Lance JW. The control of muscle tone, reflexes and movement: Robert Wartenberg lecture. Neurology 1980; 30: 1303–1313.
7. D'Amico JM, Condliffe EG, Martins KJ, et al. Recovery of neuronal and network excitability after spinal cord injury and implications for spasticity. Front Integr Neurosci 2014; 8: 36.
8. Nielsen JB, Crone C, Hultborn H. The spinal pathophysiology of spasticity: from a basic science point of view. Acta Physiol (Oxf) 2007; 189: 171–180.
9. Ashworth B. Preliminary trial of carisoprodol in multiple sclerosis. Practitioner 1964; 192: 540–542.
10. Benz EN, Hornby TG, Bode RK, et al. A physiologically based clinical measure for spastic reflexes in spinal cord injury. Arch Phys Med Rehabil 2005; 86: 52–59.
11. Biering-Sørensen F, Nielsen JB, Klinge K. Spasticity-assessment: a review. Spinal Cord 2006; 44: 708–722.
12. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987; 67: 206–207.
13. Mishra C, Ganesh GS. Inter-rater reliability of modified modified Ashworth scale in the assessment of plantar flexor muscle spasticity in patients with spinal cord injury. Physiother Res Int 2014; 19: 231–237.
14. Hsieh JT, Wolfe DL, Miller WC, et al. Spasticity outcome measures in spinal cord injury: psychometric properties and clinical utility. Spinal Cord 2008; 46: 86–95.
15. Adams MM, Martin Ginis KA, Hicks AL. The Spinal Cord Injury Spasticity Evaluation Tool: development and evaluation. Arch Phys Med Rehabil 2007; 88: 1185–1192.
16. Akman MN, Bengi R, Karatas M, et al. Assessment of spasticity using isokinetic dynamometry in patients with spinal cord injury. Spinal Cord 1999; 37: 638–643.
17. Franzoi AC, Castro C, Cardone C. Isokinetic assessment of spasticity in subjects with traumatic spinal cord injury (ASIA A). Spinal Cord 1999; 37: 416–420.
18. Grippo A, Carrai R, Hawamdeh Z, et al. Biomechanical and electromyographic assessment of spastic hypertonus in motor complete traumatic spinal cord-injured individuals. Spinal Cord 2011; 49: 142–148.
19. Lechner HE, Frotzler A, Eser P. Relationship between self-and clinically rated spasticity in spinal cord injury. Arch Phys Med Rehabil 2006; 87: 15–19.
20. Penn RD. Intrathecal baclofen for spasticity of spinal origin: seven years of experience. J Neurosurg 1992; 77: 236–240.
21. Sköld C. Spasticity in spinal cord injury: self- and clinically rated intrinsic fluctuations and intervention-induced changes. Arch Phys Med Rehabil 2000; 81: 144–149.
22. Sehgal N, McGuire JR. Beyond Ashworth. Electrophysiologic quantification of spasticity. Phys Med Rehabil Clin N Am 1998; 9: 949–979.
23. Voerman GE, Gregoric M, Hermens HJ. Neurophysiological methods for the assessment of spasticity: the Hoffmann reflex, the tendon reflex and the stretch reflex. Disabil Rehabil 2005; 27: 33–68.
24. Priebe MM, Sherwood AM, Thornby JI, et al. Clinical assessment of spasticity in spinal cord injury: a multidimensional problem. Arch Phys Med Rehabil 1996; 77: 713–716.
25. Johnson RL, Gerhart KA, McCray J, et al. Secondary conditions following spinal cord injury in a population-based sample. Spinal Cord 1998; 36: 45–50.
26. Milinis K, Young CA. Trajectories of Outcome in Neurological Conditions (TONiC) study. Systematic review of the influence of spasticity on quality of life in adults with chronic neurological conditions. Disabil Rehabil 2015; 29: 1–11.
27. Noonan VK, Kopec JA, Zhang H, et al. Impact of associated conditions resulting from spinal cord injury on health status and quality of life in people with traumatic central cord syndrome. Arch Phys Med Rehabil 2008; 89: 1074–1082.
28. Westerkam D, Saunders LL, Krause JS. Associations of spasticity and life satisfaction after spinal cord injury. Spinal Cord 2011; 49: 990–994.
29. Cardenas DD, Ditunno JF, Graziani V, et al. Two phase 3, multicenter, randomized, placebo-controlled clinical trials of fampridine-SR for treatment of spasticity in chronic spinal cord injury. Spinal Cord 2014; 52: 70–76.
30. Handa Y, Naito A, Watanabe S, et al. Functional recovery of locomotor behavior in the adult spinal dog. Tohoku J Exp Med 1986; 148: 373–384.
31. Koopmans GC, Deumens R, Honig WM, et al. The assessment of locomotor function in spinal cord injured rats: the importance of objective analysis of coordination. J Neurotrauma 2005; 22: 214–225.
32. Olby NJ, De Risio L, Munana KR, et al. Development of a functional scoring system in dogs with acute spinal cord injuries. Am J Vet Res 2001; 62: 1624–1628.
33. Olby NJ, Lim JH, Babb K, et al. Gait scoring in dogs with thoracolumbar spinal cord injuries when walking on a treadmill. BMC Vet Res 2014; 10: 58.
34. Olby NJ, Muguet-Chanoit AC, Lim JH, et al. A placebo-controlled, prospective, randomized clinical trial of polyethylene glycol and methylprednisolone sodium succinate in dogs with intervertebral disk herniation. J Vet Intern Med 2016; 30: 206–214.
35. Cook KF, Teal CR, Engebretson JC, et al. Development and validation of Patient Reported Impact of Spasticity Measure (PRISM). J Rehabil Res Dev 2007; 44: 363–371.
36. De Lahunta A, Glass E. Upper motor neuron. In: De Lahunta A, Glass E, eds. Veterinary neuroanatomy and clinical neurology. 3rd ed. St Louis: Saunders Elsevier, 2009; 192–220.
37. Dewey CW. Functional and dysfunctional neuroanatomy: the key to lesion localization. In: Dewey CW, ed. A practical guide to canine and feline neurology. 2nd ed. Ames, Iowa: Wiley-Blackwell, 2008; 17–52.
38. King AS. Pyramidal system and extrapyramidal system. In: King AS, ed. Physiological and clinical anatomy of domestic mammals: central nervous system Vol 1. Oxford, England: Oxford University Press, 1987; 141–157.
39. Schieber MH. Chapter 2. Comparative anatomy and physiology of the corticospinal system. Handb Clin Neurol 2007; 82: 15–37.
40. Fink KL, Cafferty WB. Reorganization of intact descending motor circuits to replace lost connections after injury. Neurotherapeutics 2016; 13: 370–381.
41. Han Q, Cao C, Ding Y, et al. Plasticity of motor network and function in the absence of corticospinal projection. Exp Neurol 2015; 267: 194–208.
42. de Leon RD, Tamaki H, Hodgson JA, et al. Hindlimb locomotor and postural training modulates glycinergic inhibition in the spinal cord of the adult spinal cat. J Neurophysiol 1999; 82: 359–369.
