Reevaluation of Predictive Factors for Complete Recovery in Dogs With Nonambulatory Tetraparesis Secondary to Cervical Disk Herniation
The vast majority of dogs with cervical disk herniation experience cervical pain and only mild motor deficits; therefore, not much is known about the factors that predict recovery in dogs with nonambulatory tetraparesis (NAT) secondary to cervical disk herniation. In this retrospective study, we tested the hypothesis that two previously reported prognostic factors, site of disk herniation and severity of neurological deficits, are useful predictors of complete recovery. Overall, 20 (62%) of 32 dogs with cervical disk herniation-associated NAT had complete recovery. Site of disk herniation was not a significant predictor of complete recovery; dogs with high cervical lesions (C2 to C3, C3 to C4) did not have a higher likelihood of complete recovery than other dogs. Likewise, severity of neurological deficits (i.e., intact voluntary motor function versus absent voluntary motor function) was not a significant predictor of complete recovery. Using stepwise logistic regression, two significant predictors of complete recovery were identified. Small dogs (δ15 kg body weight) were six times more likely to achieve complete recovery than larger dogs. Dogs that regained the ability to walk within 96 hours after surgery were seven times more likely to completely recover than dogs not walking 96 hours after surgery. We conclude that neither the site of disk herniation nor severity of neurological deficits assists the clinician in predicting postoperative outcome in dogs with cervical disk herniation-associated NAT. Reliable preoperative predictors of complete recovery are needed to advance current diagnostic and treatment protocols to improve overall prognosis.
Introduction
In the field of veterinary neurology, prognostic factors for complete recovery of dogs after surgical treatment of thoracolumbar disk disease are well established. Reliable predictors include severity of neurological deficits1–3 and duration of neurological signs prior to loss of ambulation.4 In dogs with thoracolumbar disk disease, site of herniation (upper motor neuron lesion versus lower motor neuron lesion) does not influence prognosis after decompressive surgery.3,5
In contrast, little is known regarding prognostic factors for surgically treated dogs with nonambulatory tetraparesis (NAT) secondary to cervical disk herniation. Only one previous case series,6 published almost 20 years ago, has focused on this understudied subset of dogs in which cervical disk herniation is accompanied by profound motor deficits. In that study, lower cervical (lower motor neuron) lesions and severity of neurological deficits were associated with a poorer prognosis for complete recovery.6 Neither duration of signs prior to NAT nor duration of NAT prior to surgery was a significant predictor of outcome.6
The purpose of this study was to reevaluate factors that predict complete recovery in surgically treated dogs with cervical disk herniation-associated NAT. Herein, we report the results of the first multivariate analysis of prognostic factors in 32 dogs.
Materials and Methods
Case Selection
The Veterinary Medical Data Base was searched for dogs admitted to the Purdue University Veterinary Teaching Hospital between January 1, 1990, and December 31, 2005, that had a diagnosis of cervical disk disease, were unable to walk, and underwent decompressive spinal surgery. In each case, myelogram showed an extradural compressive lesion centered over the disk space. Diagnosis of cervical disk herniation was confirmed at the time of surgery by the identification of disk material within the vertebral canal. For each dog, information was tabulated from medical records and telephone contact with owners and referring veterinarians. Information included the following: signalment; history of trauma; site of disk herniation; duration of signs prior to NAT; duration of NAT prior to surgery; degree of sensory and motor deficits in the thoracic and pelvic limbs; date and type of decompressive surgery; corticosteroid treatment; time to reach ambulatory status; and long-term outcome, including recurrence of clinical signs. To determine the percentage of dogs that achieved complete recovery, cases without complete recovery were excluded from analysis if follow-up of at least 12 months after surgery could not be obtained. Doberman pinschers were also excluded, because cervical disk disease in this breed is a complex syndrome associated with vertebral instability.7,8 Thirty-two dogs with cervical disk herniation-associated NAT that satisfied the inclusion criteria are the subject of this report.
Analysis of Potential Prognostic Factors
Degree of neurological recovery was recorded for each dog. The primary endpoint of this study, complete recovery, was defined as resolution of cervical pain and regaining the ability to walk normally. Cases were analyzed to determine if the following potential prognostic factors were predictive of complete recovery: body weight (≤15 kg versus >15 kg); severity of preoperative neurological deficits (e.g., absent voluntary motor function in thoracic limbs, absent voluntary motor function in pelvic limbs, intact voluntary motor function); duration of signs prior to NAT (tertiles); duration of NAT prior to surgery (tertiles); site of disk herniation (“high” lesions [intervertebral spaces C2 to C3 and C3 to C4] versus “low” lesions [intervertebral spaces C4 to C5, C5 to C6, or C6 to C7]); history of trauma (yes versus no); time to reach ambulatory status after surgery (≤96 hours versus >96 hours); and chronological age (≤8.8 years versus >8.8 years). To obtain a closer look at the relationship between age and outcome, the likelihood of complete recovery was compared in the youngest and oldest dogs after adjusting the chronological age of each dog to physiological age, a process which takes into account breed-specific differences in life expectancy.9 The association between date of surgery (three categories) and complete recovery was also evaluated to rule out a significant cohort effect.
Data were analyzed using SPSS (version 15.0) for Windows and SAS System (version 9.1) for Windows. Odds ratios (ORs) were calculated using unconditional logistic regression analysis to identify which factors predicted the likelihood of complete recovery. Odds ratios were considered significant if 95% confidence intervals (CIs) did not include 1.0. Stepwise multivariate logistic regression analysis was used to evaluate the influence of additional variables on those predictors found to be significant at P<0.10 in univariate analysis. Two-by-two tables were constructed, and chi-square analysis was used to evaluate the association between site of herniation and severity of neurological deficits. Fisher’s exact test was used to compare the likelihood of complete recovery in dachshunds versus other breeds. For chi-square and Fisher’s exact tests, P values <0.05 were considered significant.
Results
Clinical Features of Dogs With Cervical Disk Herniation-associated NAT
The clinical features of 32 dogs with NAT secondary to cervical disk herniation are summarized in Table 1. Median age at surgery was 8.8 years (range 3.7 to 15.0 years). Median body weight was 15 kg (range 4 to 50 kg). The dachshund was the most commonly affected breed (n=6). The most common site of disk herniation was the C5 to C6 intervertebral space (n=9 dogs), accounting for 28% of dogs in this series. Twelve (38%) of 32 dogs had high cervical lesions (upper motor neuron, C2 to C3 and C3 to C4 intervertebral spaces), whereas 20 (62%) dogs had low cervical lesions. One dog (case no. 2) had disk herniation at two sites.
All dogs had preoperative thoracic limb and pelvic limb neurological deficits severe enough to render them unable to walk. In three dogs, medical records did not specify the presence or absence of voluntary motor function. Voluntary motor function was intact in 21 (72%) of 29 dogs. Six dogs were tetraplegic, whereas two dogs had absent forelimb voluntary motor function. Deep pain perception in the thoracic limbs was absent in only one dog (case no. 11); all dogs had intact deep pain perception in the pelvic limbs. Fifteen dogs had some degree of neck pain noted in the medical record.
Six (19%) dogs had a history of trauma immediately preceding their neurological signs. Median duration of neurological signs prior to NAT was 24 hours (range 0 hours to 6 years). Median duration of NAT prior to surgery was 3 days (range <24 hours to 1 month). Dogs were treated with ventral slot decompression (n=29) or dorsal laminectomy (n=3). Twenty-seven dogs received intraoperative methyl-prednisolone succinate at a dose of 30 mg/kg body weight. Twenty-five dogs received corticosteroids for a period of days to weeks prior to surgery.
Functional Outcome in Dogs With Cervical Disk Herniation-associated NAT
Overall, 20 (62%) of 32 dogs with NAT secondary to cervical disk herniation had complete recoveries. Median duration of follow-up for this group was 51 months (range 5 to 111 months). Twelve dogs did not achieve complete recovery. Nine (28%) dogs in this series became ambulatory but had residual paresis and/or ataxia. Median duration of follow- up for this group was 47 months (range 18 to 127 months). Median time to reach ambulatory status after surgery was 6 days.
Three (9%) dogs never walked after surgery. One dog (case no. 32) was euthanized on the 5th postoperative day because of progressive dyspnea. One dog (case no. 30) was euthanized 1 month after surgery for failure to ambulate. One dog (case no. 31) was euthanized 2 months after surgery because of renal failure. None of these dogs underwent necropsy.
Overall, recurrence of signs consistent with intervertebral disk herniation occurred in three of 20 dogs that had complete recoveries. Recurrence of intervertebral disk herniation was documented in one dog and suspected in two other dogs. One dog (case no. 3) had two episodes of severe neck pain that responded to corticosteroids and cage rest. One dog (case no. 13) developed NAT 19 months after dorsal laminectomy; ventral slot decompression was performed at the same site as the initial surgery, and a large amount of fibrous material grossly consistent with disk material was removed from the vertebral canal. One dog (case no. 16) had recurrent NAT 30 months after surgery and was euthanized. Two (22%) of nine dogs with residual neurological deficits were euthanized because of recurrent NAT approximately 2 years after surgery (case nos. 23, 26). None of these dogs underwent necropsy.
Predictive Factors for Complete Recovery in Dogs With Cervical Disk Herniation-associated NAT
The association between possible prognostic factors and complete recovery was analyzed in 32 dogs [Table 2]. Neither the site of disk herniation nor severity of neurological deficits were significant predictors of complete recovery in dogs with cervical disk herniation-associated NAT. Six (50%) of 12 dogs with high cervical lesions had complete recovery, whereas 14 (70%) of 20 dogs with low cervical lesions had complete recovery (P=0.26). Five (83%) of six tetraplegic dogs and both dogs with absent voluntary motor function in the thoracic limbs had complete recovery. The one dog with absent deep pain perception in the thoracic limbs made a complete recovery.
In dogs with cervical disk herniation-associated NAT, two significant predictors of complete recovery were identified: small body size and regaining ambulatory status within 96 hours after surgery. Univariate analysis showed small dogs (≤15 kg body weight) were more than five times more likely to have complete recovery than larger dogs (OR, 95% CI = 5.6, 1.3 to 27.5). Six of six dachshunds had complete recovery, whereas 14 of 26 other breeds had complete recovery (P=0.06). Incidence of complete recovery was not different between dachshunds and other small breeds (P=0.25, Fisher’s exact test).
In univariate analysis, dogs that regained the ability to walk within 96 hours after surgery were more than six times more likely to have complete recovery, compared with dogs that remained nonambulatory at 96 hours (OR, 95% CI = 6.1, 1.1 to 35.3). Eleven (85%) of 13 dogs that regained ambulatory status within 96 hours after surgery went on to make a complete recovery, whereas only nine (47%) of 19 dogs that were still unable to walk at 96 hours completely recovered.
In stepwise logistic regression analysis, the relationship strengthened between small body size and complete recovery (OR, 95% CI = 6.8, 1.1 to 39.9) and between regaining ambulatory status within 96 hours after surgery and complete recovery (OR, 95% CI = 7.5, 1.1 to 51.6). No other factors, including chronological age, physiological age, history of trauma, duration of signs prior to NAT, duration of NAT prior to surgery, or date of surgery were significant predictors of complete recovery.
Discussion
Whereas the prognostic factors for surgically treated dogs with thoracolumbar disk herniation are well established,1–4 not much is known about which factors predict complete recovery in dogs with NAT secondary to cervical disk herniation.6,10–17 Two prognostic factors, site of disk herniation and severity of neurological deficits, are stated in veterinary textbooks;18,19 but these factors are based upon a single study of 12 cases.6 This prompted us to reevaluate prognostic factors in this subset of dogs having cervical disk herniation and severe motor deficits. From this first multivariate analysis of prognostic factors in 32 dogs with NAT, we reject the notion that site of disk herniation and severity of neurological deficits are useful predictors of outcome. Instead, our results suggest that small body size and regaining ambulatory status within 96 hours postoperatively are the strongest predictors of complete recovery in dogs with cervical disk herniation-associated NAT.
We tested the hypothesis that high cervical lesions (C2 to C3, C3 to C4) are associated with a more favorable prognosis than lower cervical lesions (C4 to C5, C5 to C6, C6 to C7). This association was suggested by a previous report in which three of three dogs with high cervical lesions had complete recoveries, whereas only five (56%) of nine dogs with lower cervical lesions had complete recoveries.6 In the present study, however, site of herniation was not a predictor of complete recovery. Dogs with lower cervical lesions were not less likely to have complete recoveries. In this report, lower cervical lesions were associated with an increased likelihood of recovery, but this difference was not statistically significant. The finding that site of herniation fails to impact prognosis in dogs with cervical disk herniation- associated NAT mirrors the senior author’s observations of surgically treated dogs with thoracolumbar disk herniation.5 Although counterintuitive, it seems that dogs with lower motor neuron lesions secondary to disk herniation do not have a poorer prognosis than dogs with upper motor neuron lesions.
Severity of neurological deficits is considered a robust predictor of outcome in surgically treated dogs with thoracolumbar disk herniation.1 In contrast, in dogs with NAT secondary to cervical disk herniation, we found no significant association between severity of neurological deficits and complete recovery. Absence of voluntary motor function in the thoracic or pelvic limbs was not a reliable predictor of poor outcome; 83% of tetraplegic dogs recovered. In a previous report,6 two of 12 dogs with NAT had absent deep pain perception in the thoracic limbs; both failed to achieve complete recovery. In the present series, however, the only dog that had sensory deficits had complete recovery. Our results indicate that fewer than one-third of dogs that are nonambulatory secondary to cervical disk herniation experience loss of voluntary motor function; sensory deficits are encountered even less frequently. While the hypothesis that complete sensorimotor loss portends a poor prognosis in these dogs remains untested, we advise clinicians to refrain from using severity of neurological deficits as a proxy for poor postoperative outcome in dogs with cervical disk herniation-associated NAT.
In contrast to the null findings for site of disk herniation and severity of neurological deficits, small body size was found to be a significant predictor of complete recovery. Since the 1989 report,6 additional data probing the association between signalment and complete recovery in dogs with cervical disk herniation-associated NAT are limited to eight new cases.10,11 All weighed ≤15 kg, and seven (86%) of eight dogs had complete recovery. Twenty-three of 23 small, tetraparetic dogs that were reported as nonambulatory by Cherrone et al17 became ambulatory after surgical treatment of cervical disk herniation. The likelihood of these dogs to completely recover (i.e., walk normally) was not reported. The reason small body size confers a more than five-fold advantage for complete recovery in our series is not clear. Small dogs with NAT were more likely to have more severe preoperative neurological deficits. Seven (47%) of 15 dogs weighing ≤15 kg had absent voluntary motor function, whereas only one (7%) of 14 larger dogs had absent voluntary motor function (P=0.04).
Dachshunds accounted for almost 40% of small dogs, and each of the six dachshunds in our series had complete recovery. This 100% complete recovery rate for dachshunds far exceeds the 50% complete recovery rate for non-dachshunds in this series and the 58% complete recovery rate reported previously for all breeds with NAT.6 Data from dachshunds reported in the literature6,10,11,14 are limited to only six dogs; four achieved complete recovery. It is possible that anatomical factors make this breed more likely to recover completely. Deserving further study are the vertebral canal:spinal cord ratio and the nature of disk herniations in dachshunds that experience severe motor deficits.
In the dogs reported here, the most powerful predictor of complete recovery was regaining ambulatory status within 96 hours of surgery. Dogs that walked within 96 hours were more than six times more likely to have complete recovery than dogs that were still nonambulatory at 96 hours after surgery. Eighty-five percent of dogs that regained ambulatory status at 96 hours went on to complete recovery, whereas the majority (53%) of dogs that were still nonambulatory at 96 hours failed to completely recover. Many dogs with NAT secondary to cervical disk herniation regain ambulatory status rapidly; this is consistent with previous observations.6 Our findings emphasize, however, that complete recovery is also possible (albeit less likely) in dogs that do not rapidly regain the ability to walk.
Conclusion
Nonambulatory tetraparesis is an infrequent manifestation of cervical disk herniation in dogs. The prognosis for affected dogs after surgical treatment is variable; only about 60% of dogs experience complete recovery. Neither site of disk herniation nor severity of preoperative neurological deficits assists the clinician in predicting outcome. Although dachshunds and other small breeds may have a higher likelihood of favorable outcome, regaining the ability to walk within 96 hours after surgery is the strongest predictor of complete recovery. Future studies should focus on identifying reliable preoperative factors associated with poor outcome in this subset of dogs with cervical disk herniation. Success in this effort may guide the development of new diagnostic and treatment protocols that will improve overall prognosis.
