Association between anesthesia duration and outcome in dogs with surgically treated acute severe spinal cord injury caused by thoracolumbar intervertebral disk herniation

Abstract Background Retrospective research recently identified a possible relationship between duration of surgery and outcome in severely affected dogs treated surgically for acute thoracolumbar intervertebral disk herniation (TL‐IVDH). Hypothesis That increased duration of surgery is associated with poorer outcome in dogs with absent pain perception treated surgically for TL‐IVDH. Animals Two hundred ninety‐seven paraplegic dogs with absent pain perception surgically treated for acute TL‐IVDH. Methods Retrospective cohort study. Medical records of 5 institutions were reviewed. Inclusion criteria were paraplegia with absence of pain perception, surgical treatment of TL‐IVDH, and 1‐year postoperative outcome (ambulatory: yes or no). Canine data, outcome, and surgery and total anesthesia duration were retrieved. Results In this study, 183/297 (61.6%) dogs were ambulatory within 1 year, 114 (38.4%) dogs failed to recover, including 74 dogs (24.9%) euthanized because of progressive myelomalacia. Median anesthesia duration in dogs that regained ambulation within 1 year of surgery (4.0 hours, interquartile range [IQR] 3.2‐5.1) was significantly shorter than those that did not (4.5 hours, IQR 3.7‐5.6, P = .01). Multivariable logistic regression demonstrated a significant negative association between both duration of surgery and total anesthesia time and ambulation at 1 year when controlling for body weight and number of disk spaces operated on. Conclusions and Clinical Importance Findings support a negative association between increased duration of anesthesia and outcome in this group of dogs. However, the retrospective nature of the data does not imply a causal relationship.

duration of surgery and total anesthesia time and ambulation at 1 year when controlling for body weight and number of disk spaces operated on.
Conclusions and Clinical Importance: Findings support a negative association between increased duration of anesthesia and outcome in this group of dogs. However, the retrospective nature of the data does not imply a causal relationship.
canine, extrusion, hemilaminectomy, prognosis, surgery 1 | INTRODUCTION Acute thoracolumbar intervertebral disk herniation (IVDH) is the leading cause of acute spinal cord injury (SCI) in dogs. 1 Spinal cord injury after acute IVDH occurs through a combination of contusive and compressive insults, leading to a cascade of progressive secondary cellular and biochemical injury pathways. [1][2][3] Treatment can be either medical or surgical, with the decision typically based on the severity of presenting clinical signs. 4,5 The most severely affected dogs are those with paraplegia and absent pain perception. Dogs presenting with this degree of severity typically undergo surgical treatment by hemilaminectomy in order to decompress the affected spinal cord. 5,6 Successful outcome rates for surgical treatment of dogs with absent pain perception secondary to thoracolumbar IVDH range from 25% to 76%, compared to close to 100% for less severely affected dogs. 4,[6][7][8][9] Whereas the variability of recovery within the group of dogs with absent pain perception can be ascribed in part to differences in initial injury severity captured by measurement of serum concentrations of glial fibrillary acidic protein and phosphorylated neurofilament heavy chain at time of injury, [10][11][12] as well as extent of injury seen on magnetic resonance imaging (MRI), 7 it is important to examine factors associated with management of these cases that could influence their ultimate outcome.
Of the many factors that contribute to the secondary injury cascade initiated by SCI, spinal cord perfusion appears to be one of the most important variables. 13,14 Indeed, outcome after experimental acute SCI correlates with the perfusion of the injured spinal cord. 2 24,25 increasing total duration of anesthesia and surgery might be negatively associated with outcome in the most severely affected dogs, those with absent pain perception. 9 The recently established Canine Spinal Cord Injury Consortium (CANSORT-SCI) was created with the aim of facilitating large-scale research into acute SCI in dogs, through collaboration between multiple international institutions. 26,27 This initiative has provided an opportunity to investigate hypotheses generated by exploratory research such as this in a large, diverse population of dogs across different institutions and countries.
The aim of this study was to investigate whether there was an association between outcome and anesthesia duration using the CANSORT-SCI to provide access to a large population of severely affected dogs surgically treated for thoracolumbar IVDH at multiple international institutions. 26 Our hypothesis was that there would be a negative association between duration of anesthesia and outcome in dogs with absent pain perception treated surgically for thoracolumbar IVDH.  26 Inclusion criteria were dogs that had undergone surgical treatment by hemilaminectomy for thoracolumbar IVDH; paraplegic with absent pain perception in both pelvic limbs and tail before surgery; available anesthesia duration records (total general anesthesia duration and surgery duration); with 1-year ambulatory status follow-up recorded.

| Study design and animals
Exclusion criteria were incomplete anesthesia duration records or lack of follow-up data, with dogs euthanized owing to signs of progressive myelomalacia (PMM) postoperatively included as an unsuccessful (nonambulatory) outcome.
Case records were retrospectively analyzed for descriptive dog (breed, age, body weight, and sex) and clinical data (imaging modality used, surgery site, and number of intervertebral disk [IVD] spaces operated on).
Duration of surgery (first incision to last suture) and total duration of anesthesia (administration of induction agent to extubation) were also retrieved from the medical records. Outcome was determined according to ambula-   In dogs that regained ambulatory function within 1 year postoperatively, total duration of anesthesia was significantly shorter, with a median of 4.0 hours (minimum 1.5, maximum 8.0, IQR 3.2-5.1), compared to 4.5 hours (minimum 1.9, maximum 10.9, IQR 3.7-5.6) in those that did not (P = .01; Figure 1). Median duration of surgery in dogs that regained ambulatory function within 1 year was 2.1 hours (minimum 0.7, maximum 6.3, IQR 1.5-2.8), while in dogs that did not recover ambulation it was 2.3 hours (minimum 0.7, maximum 7.8, IQR 1.6-3.1).

| Outcome
F I G U R E 1 Dot plots showing the distribution of total anesthesia duration (Total general anesthesia (GA) duration) for 297 dogs with absent pain perception that were ambulatory (n = 183) or nonambulatory (n = 114) within 1 year of underdoing surgical treatment for acute thoracolumbar intervertebral disk herniation. Central line = median, whiskers = interquartile range. *Statistically significant difference (P = .010) F I G U R E 2 Dot plots showing the distribution of surgery duration for 297 dogs with absent pain perception that were ambulatory (n = 183) or nonambulatory (n = 114) within 1 year of underdoing surgical treatment for acute thoracolumbar intervertebral disk herniation. Central line = median, whiskers = interquartile range. The difference in surgery duration between outcome groups was not statistically significant (P = .057) The duration of surgery was not significantly different between the 2 groups on bivariate analysis (P = .06; Figure 2). The median total duration of anesthesia for dogs that were euthanized postoperatively for presumed PMM was 4.3 hours (minimum 1.9, maximum 8.5, IQR 3.7-5.5) and the median duration of surgery was 2.3 hours (minimum

| DISCUSSION
The findings of this study provide further evidence of a negative association between increased duration of anesthesia and recovery of ambulation in severely affected dogs surgically treated for thoracolumbar IVDH. We were able to explore and further investigate the findings of recent retrospective research using data from a large cohort of dogs with absent pain perception recruited through the international CANSORT-SCI network of referral hospitals. 26,27 Although this associa-  (Tables 1 and 3). This provides further support for the findings of our previous retrospective study, 9  T A B L E 4 Multiple logistic regression model including surgery duration, body weight, and size of hemilaminectomy as predictors of ambulatory status 1 year postoperatively among 223 dogs with absent pain perception that underwent surgical treatment for acute thoracolumbar intervertebral disk herniation-excluding cases euthanized for presumed progressive myelomalacia The retrospective nature of this study led to some expected limitations including missing data points and exclusion of several cases owing to a lack of outcome information. Because of the nature of the multicenter data collection, long-term outcome was restricted to ambulatory status without confirmed pain perception status, meaning that some of the dogs in the recovery group could have developed "spinal walking." Given the evidence that dogs that recover ambulation without pain perception have intact trans-lesion connections and smaller lesions than dogs that do not recover ambulation, our emphasis was on regaining the ability to ambulate regardless of pain perception status in the long term as the primary outcome measure for success in such cases. [38][39][40] We were also not able to access detailed anesthesia protocols including specific medications that were used between institutions. However, the use of the CANSORT-SCI network allowed the collection of data from several institutions, helping to counteract these limitations. As dogs were treated at 5 different institutions, there might have been variations in surgical technique, anesthetic protocols, and management between centers.
In conclusion, the findings of this study suggest that an increased duration of anesthesia is negatively associated with outcome in paraplegic dogs with absent pain perception undergoing surgical treatment for thoracolumbar IVDH. Although further investigation is required to determine if this is a causal relationship, these results suggest that markedly prolonged anesthesia negatively might affect recovery in severely affected dogs with thoracolumbar IVDH.