Intervertebral disk herniation is one of the most common causes of spinal cord dysfunction in dogs, and small breeds and chondrodystrophic dogs are predisposed.1,2 The disease can be categorized on the basis of whether the herniation involves disk extrusion or protrusion. Intervertebral disk extrusion is characterized by complete rupture of the annulus fibrosus and subsequent displacement of the nucleus pulposus in the vertebral canal, whereas intervertebral disk protrusion is characterized by a degenerative process of the annulus fibrosus that can cause a focal and gradual extension of the annulus into the vertebral canal.3,4 In most (66% to 86%) affected dogs, the thoracolumbar portion of the vertebral column is affected.5 On the other hand, cervical disk herniation has been reported in approximately 15% of dogs with IVDH.1,6
Magnetic resonance imaging is widely considered the most accurate imaging modality for assessment of dogs with suspected IVDH, with a reported sensitivity of 98.5% in the diagnosis of IVDH7 and accuracy of 100% in the identification of site and side of the lesion.8 Moreover, MRI allows detection of intramedullary signal changes related to spinal cord inflammation, edema, hemorrhage, myelomalacia, and gliosis.9,10
Decompressive surgery is often required to remove the extruded disk material causing SCC.3,11,12 Indeed, immediate surgical decompression is frequently successful in improving ambulatory status and spinal hyperesthesia in dogs with cervical and thoracolumbar IVDH.3,13,14 Adverse events have been reported for 9.9% of decompressive surgeries to treat cervical IVDH and 1.0% to 7.7% for surgeries to treat thoracolumbar IVDH.15,16 Cervical and thoracolumbar decompressive surgeries are often not effective in completely removing the extruded material, and residual SCC is a reported cause of the recurrence or persistence of clinical signs.11,16–18 Nevertheless, follow-up imaging is not routinely performed, and only a few studies18–20 have been reported regarding the potential usefulness of CT and MRI in the postoperative assessment of SCD.
The main purpose of the study reported here was to assess agreement between the surgeon's perception of the effectiveness of SCD and postoperative MRI findings for dogs undergoing decompressive surgery for IVDE. Other objectives were to identify preoperative factors that might be associated with postoperative MRI findings of satisfactory SCD, determine whether satisfactory SCD as identified via postoperative MRI had a stronger association with outcome or recovery time than satisfactory decompression as perceived by the surgeon, and determine whether spinal cord features in pre- and postoperative MRI images were associated with surgical outcome.
Materials and Methods
Animals
A prospective study was conducted by which all dogs with cervical or thoracolumbar IVDH admitted to the Clinica Neurologica Veterinaria NVA in Milan, Italy, from January 2015 to December 2016 were considered for inclusion. This period was chosen because we anticipated that it would provide a sufficient number of dogs; no sample size calculation was performed. Only dogs with single compressive cervical or thoracolumbar IVDE that received surgical treatment were considered for inclusion in the study. All procedures were performed with the owner's written consent. An approval from an institutional animal care and use committee was not sought because all the procedures performed, including pre- and postoperative MRI, were considered routine at our clinic for patients with IVDH, as supported by the literature.7,11,17,19
All initially included dogs underwent complete physical and neurologic examinations, preoperative hematologic testing, radiographic evaluation of the vertebral column, pre- and postoperative MRI, and decompressive surgery. The diagnosis of IVDH was made by identifying extradural compression of the spinal cord at the level of an intervertebral disk space, narrowing of the intervertebral disk space, and evidence of degeneration of the intervertebral disk on preoperative MRI images.6 Dogs were excluded if findings of physical examination and hematologic testing suggested they had concurrent disease.13,14 Dogs were also excluded if their neurologic signs were not limited to the spinal cord and vertebral column or they had radiographic findings of diskospondylitis, neoplasia, traumatic lesions,2,13,14 preoperative MRI findings of multiple compressive IVDHs, or intraoperatively identified intervertebral disk protrusion. Moreover, dogs were excluded if no follow-up data were collected after surgery and discharge from the clinic.
Data collection
Age, sex, breed, body weight, and body condition score (9-point scale, with 5 representing ideal, values < 5 representing less than ideal, and values > 5 representing more than ideal21) were recorded for each dog. Breeds were further classified as chondrodystrophic (Dachshund, Beagle, Shih Tzu, Pekingese, English Bulldog, French Bulldog, Pug, Miniature Poodle, Cocker Spaniel, West Highland White Terrier, Welsh Corgi, Bichon Frise, and Jack Russell Terrier) and nonchondrodystrophic.1,22–26
A grading system from 1 to 5 was used to classify neurologic status, whereby a grade of 1 = signs of pain only, 2 = ambulatory tetraparesis or paraparesis, 3 = nonambulatory tetraparesis or paraparesis, 4 = tetraplegia or paraplegia but signs of deep pain, and 5 = tetraplegia or paraplegia and no signs of deep pain perception.25 Dogs that were neurologically normal and without signs of spinal pain during the postoperative period were assigned a grade of 0. Duration of clinical signs, defined as the time from onset of neurologic dysfunction to surgery, was classified as < 24 hours, 24 to 48 hours, or > 48 hours.
Preoperative MRI procedures
With dogs anesthetized, MRI of the cervical or thoracolumbar portion of the vertebral column was performed by use of a 0.3-T scannera and human knee coil or flexible body coil as indicated by body conformation. The MRI protocol included sagittal, transverse, and dorsal plane T2-weighted images and sagittal and transverse T1-weighted sequences. For sagittal spin-echo T1-weighted images, the TR was 659 milliseconds, TE was 15 milliseconds, and flip angle was 90°, and for sagittal fast spin-echo T2-weighted sequences, these settings were 2,734 milliseconds and 90°, respectively. Slice thickness for sagittal images was 4 mm, with an interval of 4.5 mm and field of view of 260 mm2. For transverse spin-echo T1-weighted sequences, the TR was 690 milliseconds, TE was 18 milliseconds, and flip angle was 90°, and for transverse fast spin-echo T2-weighted sequences, these settings were 2,937 milliseconds, 100 milliseconds, and 90°, respectively. Slice thickness for transverse images was 3 mm, with an interval of 3.5 mm and field of view of 180 mm2. Dorsal plane fast spin-echo T2-weighted images were acquired with a TR of 2,734 milliseconds, TE of 100 milliseconds, flip angle of 90°, slice thickness of 3.5 mm, interval of 4 mm, and field of view of 260 mm2.
All MRI images were evaluated by use of imaging analysis softwareb; all MRI measurements were standardized and repeated 3 times by the same reader to obtain a mean value. To evaluate the reproducibility of SCC measurements, ICCs for the 3 measurements were calculated (0 = complete absence of agreement and 1 = excellent agreement). The degree of degeneration of the affected intervertebral disk was assessed on sagittal T2-weighted images in accordance with the validated Pfirrmann 5-point grading system, whereby 1 represents an intervertebral disk with a homogeneous appearance and bright hyperintense white signal intensity and 5 represents an intervertebral disk with nonhomogeneous appearance and black hypointense signal intensity.24,27 Extruded disk material was classified as dispersed if this material spread cranially or caudally along the epidural space, losing its contact with the relative intervertebral disk space. If the extruded disk material remained just above and around the intervertebral disk space, it was classified as nondispersed.28
Sagittal T2-weighted images were used to measure horizontal and vertical extension of the extruded disk material.28 Horizontal extension was calculated as the ratio between the horizontal length of the extruded material to the length of the vertebral body (L2 for thoracolumbar IVDE and C6 for cervical IVDE). Vertical extension was calculated as the ratio between height of the extruded disk material and height of the vertebral canal.28 Transverse MRI images were used to evaluate circumferential distribution of disk herniation (classified as ventral, ventrolateral, or dorsolateral).
The degree of SCC was calculated on transverse T2-weighted images by measuring the CSA of the normal and compressed regions of the spinal cord as described elsewhere.6,29 The CSA was measured by tracing the outline of the spinal cord with an image analysis software tool, and the CSA of the spinal cord at the site of maximal compression was compared with the CSA of an unaffected region of the spinal cord immediately cranial or caudal to the site of the maximum compression (Figure 1). The degree of SCC was calculated as ([CSA of unaffected spinal cord - CSA of spinal cord under maximal compression]/CSA of unaffected spinal cord) × 100. Finally, the spinal cord was evaluated for intramedullary signal changes, and hyperintensity on T2-weighted images was recorded, if present.
Transverse T2-weighted MRI images of a 6-year-old Shih Tzu before (A) and after (B) SCD surgery to treat C3–4 disk extrusion and of a 6-year-old Dachshund before (C) and after (D) SCD surgery to treat L3–4 disk extrusion. The CSA of the spinal cord is outlined (yellow). The insets in the bottom right corner of each figure show the cross references of the transverse images.
Citation: Journal of the American Veterinary Medical Association 256, 2; 10.2460/javma.256.2.210
Surgical procedures and intraoperative evaluation
A board-certified veterinary neurologist (RL) and 2 veterinary surgeons with > 15 years of experience in spinal surgery were involved in the study (ie, 3 surgeons in total). A standard ventral slot procedure was performed for cervical disk herniations.30 Hemilaminectomies or minihemilaminectomies with a dorsolateral approach were performed for thoracolumbar disk herniations, according to the surgeon's preference and characteristics of extruded disk material (circumferential distribution and dispersion) seen in preoperative MRI images.31
The 3 surgeons were asked to evaluate the degree of surgical SCD achieved as satisfactory (residual SCC ≤ 15% in relation to the expected normal shape of the spinal cord) or unsatisfactory (residual SCC > 15% in relation to the expected normal shape of the spinal cord) on the basis of whether all visible extruded material had been removed, there was no further palpable material after probing all the aspects of the vertebral canal, and the spinal cord appeared to be in the normal position.
Postoperative MRI
Postoperative MRI was performed immediately after surgery when possible or otherwise within 7 days after surgery. The same protocol was used as for preoperative MRI. The amount of residual SCC was determined on T2-weighted transverse images through evaluation of CSA (Figure 1). The degree of SCD was considered satisfactory (normal shape of the spinal cord or residual SCC ≤ 15%) or unsatisfactory (residual SCC > 15%) as described elsewhere.20 Whether intramedullary hyperintensity was visible on T2-weighted images was also recorded.
Postoperative care and clinical outcome
Dogs were hospitalized for at least 24 hours after surgery and monitored for neurologic status and signs of pain. They were provided with cage rest and postoperative treatments that included enrofloxacinc (5 mg/kg [2.3 mg/lb], PO, q 24 h for 10 days), amoxicillin–clavulanic acidd (22 mg/kg [10 mg/lb], PO, q 12 h for 10 days), tramadole (2 mg/kg [0.9 mg/lb], PO, q 12 h for 5 days), and anti-inflammatory drugs such as carprofenf (2 mg/kg, PO, q 12 h for 7 days) or prednisoloneg (1 mg/kg [0.45 mg/lb], PO, q 12 h and subsequently tapered for 7 days). If treatment with an anti-inflammatory drug (NSAID or corticosteroid) had been initiated prior to admission to the hospital, the same medication was used subsequently.
For dogs with neurologic grades from 3 to 5, a physiotherapy program was recommended. Neurologic status was reevaluated and graded 3 to 7 days and 21 to 31 days after surgery. Dogs with recurrence of clinical signs were reexamined, and a neurologic examination was again performed; those with signs of spinal hyperesthesia or mild paresis and ataxia were managed medically.13,14,32
A follow-up telephone interview with each dog's owner was conducted 6 months after surgery. During this interview, owners were asked to reply to a series of questions regarding the presence or absence of signs of spinal pain and whether the dog was able to walk, had urinary continence, and had recurrence of signs of pain. The final outcome was considered successful if dogs with a preoperative neurologic grade of 3 to 5 recovered signs of deep pain perception, urinary continence, and the ability to walk; dogs with a preoperative neurologic grade of 2 improved at least 1 subgrade in severity of ataxia or paresis; and dogs with preoperative neurologic grade of 1 had resolution of signs of spinal pain.32 The final outcome was considered successful with recurrence if the dog had subsequent episodes of signs of spinal pain or mild paresis that resolved with exercise reduction and medical management. It was also considered unsuccessful if the dog was not able to walk unassisted, there was no change in clinical status from before surgery, or the dog was euthanized because of poor neurologic improvement. Furthermore, the final outcome was considered unsuccessful if dogs required surgical revision because postoperative MRI findings indicated the degree of SCD was unsatisfactory; those dogs were then excluded from subsequent clinical data collection (postoperative neurologic grade and recovery time data) and statistical analysis. Finally, recovery time was defined as the number of days required to achieve a successful outcome after surgery, within an evaluation period of 6 months postoperatively.
Statistical analysis
Categorical data are summarized as counts (percentages) and continuous data as mean, SD, median, and range or interquartile (25th to 75th percentile) range. The number of dogs with missing data are reported for each variable when relevant. For the analysis, 2 patient groups were created: one in which the surgeon perceived the extent of achieved SCD as satisfactory and the other in which it was perceived as unsatisfactory or uncertain. Thereafter, agreement was assessed between postoperative MRI findings and the surgeon's judgment by calculation of the Cohen κ coefficient and related 95% confidence interval, with κ values ≤ 0.20 interpreted as slight agreement, values > 0.20 and ≤ 0.40 as fair agreement, values > 0.40 and ≤ 0.60 as moderate agreement, values > 0.60 and ≤ 0.80 interpreted as substantial agreement, and values > 0.80 as near perfect agreement.
To identify potential predictors of satisfactory postoperative decompression, differences between dogs with versus without satisfactory decompression as identified via postoperative MRI were examined with respect to selected preoperative characteristics (ie, chondrodystrophic vs nonchondrodystrophic breed, body weight, neurologic grade, duration of clinical signs, region of intervertebral disk extrusion [cervical or thoracolumbar], Pfirrmann grade of disk degeneration, dispersion and extension features of extruded disk material, circumferential distribution of extruded material [ventral, ventrolateral, or dorsolateral], degree of SCC, and intramedullary hyperintensity [yes vs no]). Differences were also examined between dogs with versus without satisfactory SCD (first as assessed via postoperative MRI and then as assessed via surgeon's perception) with respect to follow-up outcome variables (postoperative neurologic grade at 3 to 7 days and 21 to 30 days after surgery and final outcome [successful vs unsuccessful or successful but with recurrence of clinical signs]) as well as pre- and postoperative intramedullary hyperintensity identified on T2-weighted MRI images (yes vs no). For both sets of analyses, the Fisher exact and Wilcoxon rank sum tests were used for categorical and continuous variables (regardless of the data distribution), respectively, given the small sample sizes. Dogs perceived by surgeons as having uncertain or unsatisfactory decompression were grouped together.
Recovery time was analyzed with the Kaplan-Meier method, which accounts for patients that were censored during follow-up, and the mean recovery time and related SD were estimated for the entire population. The same method and the log-rank test were applied to compare recovery times between dogs with and without satisfactory postoperative decompression (first as assessed via postoperative MRI and then as assessed via surgeon's perception) and between dogs with and without preoperative intramedullary hyperintensity and those with and without postoperative intramedullary hyperintensity. Values of P < 0.05 were considered significant. All analyses were performed with the aid of statistical software.h
Results
Animals and preoperative factors
Sixty-eight dogs were included in the study. Median age at IVDE diagnosis was 6 years (range, 3 to 14 years). Twenty-nine (43%) dogs were female, and 39(57%) were male. Median body weight was 7 kg (15 lb; range, 2 to 34 kg [4 to 75 lb]), and median body condition score on a 9-point scale was 5 (ideal; range, 4 to 8). Fifty-two (76%) dogs were classified as chondrodystrophic and 16 (24%) as nonchondrodystrophic.
Nineteen (28%) dogs had a preoperative neurologic grade of 1, 32 (47%) had a grade of 2, 7 (10%) had a grade of 3, 7 (10%) had a grade of 4, and 3 (4%) had a grade of 5. Given the small number of dogs with each grade, dogs were grouped according to whether they had grades of 1 or 2, 3, or 4 or 5 (reflecting ambulatory status) as previously proposed.23 The duration of clinical signs was < 24 hours for 8 (12%) dogs, 24 to 48 hours for 12 (18%) dogs, and > 48 hours for 48 (70%) dogs. The region of IVDE was thoracolumbar in 54 (79%) dogs and cervical in 14 (21%) dogs. Intervertebral disk degeneration was classified as Pfirrmann grade 1 for 1 (1%) dog, grade 2 for 3 (4%) dogs, grade 3 for 17 (25%) dogs, grade 4 for 28 (41%) dogs, and grade 5 for 19 (28%) dogs. For statistical analysis, dogs were grouped according to whether they had grade 1 (homogeneous and bright white regions on MRI), grade 2 or 3 (nonhomogeneous and gray), or grade 4 or 5 (nonhomogeneous and black) degeneration.
The median preoperative degree of SCC was 45.6% (range, 17.5% to 82.3%). The ICC value for triplicate SCC measurements was 0.96, indicating excellent reproducibility. Intramedullary hyperintensity was identified in preoperative T2-weighted MRI images of 12 (18%) dogs.
Surgical procedures and intraoperative outcomes
Ventral slot procedures were performed in 14 (21%) dogs, hemilaminectomy in 42 (62%) dogs, and minihemilaminectomy in 12 (18%) dogs. The surgeon perceived the achieved SCD as satisfactory (residual SCC ≤ 15%) in 62 (91%) dogs and as unsatisfactory in 2 (3%) dogs. For 4 (6%) dogs, the surgeon was uncertain about the degree of SCD achieved.
Postoperative MRI
Postoperative MRI was performed immediately after surgery for 51 (75%) dogs, > 24 hours after surgery for 11 (16%) dogs, > 48 hours after surgery for 1 (1%) dog, and 3, 5, and 7 days after surgery for 2 (3%), 1 (1%), and 2 (3%) dogs, respectively. The median degree of SCC as assessed via postoperative MRI was 8.8% (range, 2.3% to 47.3%), and the achieved SCD was considered satisfactory in 53 (78%) dogs. Reproducibility of triplicate SCC measurements was very high, with an ICC value of 0.98.
Intramedullary hyperintensity was detected on postoperative T2-weighted MRI images in 23 (34%) dogs, and no such hyperintensity was detected in 41 (60%) dogs. Four (6%) dogs had missing data on this variable.
The agreement between the surgeon's perception and postoperative MRI findings regarding whether satisfactory SCD had been achieved was only fair (κ = 0.40; 95% confidence interval, 0.13 to 0.67). Indeed, for 11 dogs, the surgeon's perception disagreed with postoperative MRI findings; 10 of these dogs had thoracolumbar IVDE, and 1 had cervical IVDE. In 10 dogs, SCD was perceived as satisfactory by the surgeon, but postoperative MRI findings indicated it was unsatisfactory; 2 of these dogs underwent a second surgery to remove residual disk material. Among the 4 dogs for which the surgeon was uncertain about the amount of SCD achieved, postoperative MRI findings indicated it was satisfactory for 1 dog and unsatisfactory for the remaining 3 dogs. Surgical revision was performed in 1 of these 4 dogs. All dogs that underwent a second surgery had a thoracolumbar disk extrusion with ventral circumferential distribution of extruded material. In 2 dogs, the surgeon perceived the degree of SCD as unsatisfactory, and postoperative MRI findings confirmed this perception.
Preoperative neurologic grade, region of IVDE, and circumferential distribution of disk extrusion were significantly associated with the degree of postoperative residual SCC identified by MRI. Dogs with a preoperative neurologic grade of 1 or 2 were more likely to have satisfactory SCD than were dogs with a more severe grade (Table 1). Similarly, dogs with cervical disk extrusion and dogs with ventrolateral or dorsolateral circumferential distribution of the extruded disk material were more likely to have satisfactory SCD than were dogs without these factors. Dogs with satisfactory SCD had a significantly lower degree of SCC before surgery than did dogs with unsatisfactory decompression.
Preoperative factors associated with satisfactory SCD* as assessed via postoperative MRI in 68 dogs surgically treated for cervical or thoracolumbar IVDE.
| Amount of decompression achieved | |||
|---|---|---|---|
| Factor | Unsatisfactory (n = 15) | Satisfactory (n = 53) | P value |
| Chondrodystrophic breed (No. [%]) | 0.74 | ||
| Yes | 11 (73) | 41 (77) | - |
| No | 4 (27) | 12 (23) | - |
| Body weight (kg) | 0.88 | ||
| Mean (SD) | 7.9 (4.3) | 8.7 (5.9) | - |
| Median (IQR) | 6.6 (6.0–8.7) | 6.8 (5.9–9.0) | - |
| Neurologic grade (No. [%]) | 0.01 | ||
| 1–2 | 7 (47) | 44 (83) | - |
| 3 | 3 (20) | 4 (8) | - |
| 4–5 | 5 (33) | 5 (9) | - |
| Duration of clinical signs (h; No. [%]) | 0.42 | ||
| < 24 | 3 (20) | 5 (9) | - |
| 24–48 | 3 (20) | 9 (17) | - |
| < 48 | 9 (60) | 39 (74) | - |
| Region of IVDE (No. [%]) | 0.03 | ||
| Cervical | 0 (0) | 14 (26) | - |
| Thoracolumbar | 15 (100) | 39 (74) | - |
| Grade of disk degeneration (No. [%]) | 0.38 | ||
| 1 | 0 (0) | 1 (2) | - |
| 2–3 | 2 (13) | 18 (34) | - |
| 4–5 | 13 (87) | 34 (64) | - |
| Dispersed disk material (No. [%]) | 0.23 | ||
| Yes | 3 (20) | 20 (38) | - |
| No | 12 (80) | 33 (62) | - |
| Vertical extension (%)† | 0.49 | ||
| Mean (SD) | 61 (15) | 58 (20) | - |
| Median (IQR) | 61 (45–72) | 56 (47–66) | - |
| Horizontal extension (%)† | 0.07 | ||
| Mean (SD) | 62 (32) | 81 (42) | - |
| Median (IQR) | 56 (40–65) | 70 (51–92) | - |
| Circumferential distribution of extruded material (No. [%]) | 0.02 | ||
| Ventral | 9 (60) | 12 (23) | - |
| Ventrolateral | 6 (40) | 38 (72) | - |
| Dorsolateral | 0 (0) | 3 (6) | - |
| Degree of SCC (%) | 0.0499 | ||
| Mean (SD) | 50.7 (12.8) | 42.9 (12.8) | - |
| Median (IQR) | 50.7 (41.1–60.0) | 44.6 (35.0–50.9) | - |
| Intramedullary hyperintensity on MRI images (No. [%]) | 0.90 | ||
| Yes | 2 (13) | 10 (19) | - |
| No | 12 (80) | 37 (70) | - |
| Uncertain | 1 (7) | 6 (11) | - |
For neurologic grades and grades of disk degeneration, the higher the value is, the more severe the status was.
Satisfactory SCD was defined as a residual amount of SCC ≤ 15% in relation to the expected normal shape of the spinal cord.
Values were missing for 1 dog assessed as having satisfactory decompression.
IQR = Interquartile (25th to 75th percentile) range.
Follow-up outcomes
By 3 to 7 days after surgery, 19 (28%) dogs had a neurologic grade of 0 and 36 (53%) dogs had a neurologic grade of 1 or 2. By 21 to 30 days after surgery, 33 (49%) dogs had a neurologic grade of 0 and 27 (40%) dogs had a neurologic grade of 1 or 2. For statistical analysis, dogs were grouped according to whether they had grades of 0, 1 or 2, 3, or 4 or 5. The final outcome was considered successful for 52 (76%) dogs, successful but with recurrence of clinical signs for 10 (15%) dogs, and unsuccessful for 6 (9%) dogs.
The distribution of neurologic grades at 3 to 7 days (P = 0.004) and 21 to 30 days (P = 0.001) after surgery differed significantly between dogs with versus without satisfactory SCD as indicated by postoperative MRI findings (Figure 2), indicating greater proportions of dogs with more severe neurologic statuses in the group with unsatisfactory SCD. The proportion with a successful final outcome was significantly (P = 0.005) greater for dogs with satisfactory SCD (45/53 [85%]) than for dogs with unsatisfactory decompression (7/15 [47%]). Mean recovery time was also significantly longer for dogs with unsatisfactory decompression (33.1 days; SD, 4.0 days) than for dogs with satisfactory decompression (18.2 days; SD, 1.8 days; P = 0.009; Figure 3).
Distributions of neurologic grades for dogs 3 to 7 days and 21 to 30 days after SCD surgery to treat IVDE in which postoperative MRI findings indicated that the amount of SCD achieved was satisfactory (residual SCC ≤ 15%; gray bars; n = 53) or unsatisfactory (residual SCC > 15%; white bars; 15). A grade of 0 indicates normal findings on neurologic examination and no signs of spinal pain, 1 to 2 indicates ambulatory tetraparesis or paraparesis, 3 indicates nonambulatory tetraparesis or paresis, and 4 to 5 indicates tetraplegia or paraplegia. Distributions of dogs with the various grades differed significantly (P ≤ 0.004) between groups, indicating greater proportions of dogs with more severe statuses in the group with unsatisfactory SCD.
Citation: Journal of the American Veterinary Medical Association 256, 2; 10.2460/javma.256.2.210
Kaplan-Meier curves of recovery time for dogs following SCD surgery (0 days) to treat IVDE, grouped by whether the surgery resulted in satisfactory (solid line) or unsatisfactory (dashed line) SCD as assessed via postoperative MRI findings. Recovery time was defined as the number of days required to achieve a successful outcome after surgery, within a postoperative evaluation period of 6 months. See Figure 2 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 2; 10.2460/javma.256.2.210
The distribution of neurologic grades at 21 to 30 days (but not 3 to 7 days) after surgery differed significantly (P = 0.02) between dogs with versus without satisfactory SCD as perceived by the surgeon during surgery (Figure 4), indicating greater proportions of dogs with more severe neurologic grades in the group with unsatisfactory SCD. The proportion with a successful final outcome did not differ significantly (P = 0.14) between dogs perceived as having satisfactory SCD (49/62 [79%]) and those perceived as having unsatisfactory decompression (3/6 [50%]). Dogs for which SCD was perceived as successful by the surgeon had a significantly shorter mean recovery time (19.1 days; SD, 1.7 days) than did dogs for which SCD was perceived as unsatisfactory (45.0 days; SD, 2.9 days; Figure 5).
Distributions of neurologic grades for dogs 3 to 7 days and 21 to 30 days after SCD surgery to treat IVDE for which the amount of SCD achieved as perceived by the surgeon was satisfactory (residual SCC ≤ 15%; gray bars; n = 62) or unsatisfactory or uncertain (residual SCC > 15%; white bars; 6). Distributions of dogs with the various statuses differed significantly (P = 0.02) between groups only at 21 to 30 days after surgery (and not at 3 to 7 days after surgery). See Figure 2 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 2; 10.2460/javma.256.2.210
Kaplan-Meier curves of recovery time for dogs following SCD surgery (0 days) to treat IVDE, grouped by whether the surgery resulted in satisfactory (solid line) or unsatisfactory (dashed line) SCD as perceived by the surgeon. See Figures 2 and 4 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 2; 10.2460/javma.256.2.210
The distribution of postoperative neurologic grades at 3 to 7 days (but not 21 to 30 days) after surgery differed significantly (P = 0.004) between dogs with intramedullary hyperintensity identified on preoperative MRI images, indicating that intramedullary hyperintensity was more common in dogs with a more severe neurologic status. Proportions of dogs with preoperative intramedullary hyperintensity did not differ significantly on the basis of final outcome (P > 0.99) or recovery time (P = 0.87). Proportions of dogs with intramedullary hyperintensity identified on postoperative MRI images did not differ significantly on the basis of neurologic status at 3 to 7 days (P = 0.09) or 21 to 30 days (P = 0.80) after surgery, final outcome (P = 0.37), or recovery time (P = 0.62).
Discussion
The results of the present study indicated that the surgeon's perception and postoperative MRI findings regarding the degree of SCD were only fairly correlated. In most instances in which the MRI results differed from the surgeon's perception, the degree of surgical decompression was perceived as satisfactory by the surgeon but revealed as unsatisfactory via MRI. This finding suggested that surgeons might overestimate the amount of disk material removed or underestimate the volume of disk compressing the spinal cord. However, postoperative MRI revealed satisfactory SCD most often in dogs with a less severe neurologic grade, and among all dogs included in the study, dogs with a low preoperative neurologic grade were overrepresented. Accordingly, it cannot be excluded that this situation might have influenced the agreement between the surgeon's perception and postoperative MRI findings regarding SCD.
In most instances in which MRI findings and the surgeon's perception disagreed, the degree of SCD was assessed as unsatisfactory on postoperative MRI images. The median residual amount of SCC as identified via postoperative MRI was 8.8%, and SCD was considered satisfactory by means of this modality in 53 of 68 (78%) dogs.
Little information is available regarding the use of advanced diagnostic imaging to assess residual extruded disk material in dogs following SCD surgery.11,12,18–20 Residual SCC was identified by means of postoperative CT myelography in 10 of 13 dogs undergoing ventral slot surgery in 1 study.18 In another study,11 postoperative CT revealed residual disk material in 100% (40/40) of dogs treated by hemilaminectomy. A subsequent study19 identified residual disk material via postoperative MRI in 4 of 9 dogs that underwent minihemilaminectomy.
Most surgeons rely on visual investigation and probing of the vertebral canal to establish satisfactory removal of disk material.15,32 The size of the affected dog, anatomic region of disk material, limited surgical view, and presence of concurrent hemorrhage could influence the degree of SCD achieved.2,11,33 In the present study, body weight was considered a measurement of the size of the dogs. This variable did not differ significantly between dogs with and without satisfactory SCD as indicated by postoperative MRI findings, and this lack of a difference might have been attributable to the high proportion (57/68 [84%]) of dogs with a body weight < 10 kg (22 lb).
Dogs with a severe degree of SCC before surgery in the present study were more likely to have unsatisfactory SCD than other dogs. This result could have been due to greater difficulty in completely removing extruded disk material during surgery; in patients with severe compression, the surgeon might perceive that satisfactory SCD has been achieved despite the presence of some residual disk material in the vertebral canal. Previous studies6,34 involving dogs with cervical and thoracolumbar disk extrusion have shown a pattern or a positive association between severity of presurgical neurologic grade and the degree of presurgical SCC; therefore, dogs with a severe preoperative neurologic grade may have severe preoperative SCC and could be at increased risk of unsatisfactory SCD.
No consensus exists regarding the clinical importance of residual SCC. Previous studies17,35 of dogs with thoracolumbar disk herniation treated by hemilaminectomy have shown that the residual extruded disk material is a cause of continued SCC and poor postoperative outcome. In contrast, in another study,11 all dogs reacquired voluntary ambulation and urinary control despite the presence of residual herniated disk material in the vertebral canal. However, that study did not involve investigation of a possible association between the degree of SCD and improvement of neurologic status or recurrence of clinical signs after surgery. In another study,18 no correlation was found between residual SCC and postoperative neurologic function in 17 surgically treated dogs with cervical IVDH; nevertheless, at least 1 episode of stiffness or signs of cervical pain was reported for 6 of these dogs.
In the present study, results suggested an association between the degree of SCD identified via postoperative MRI and outcome. Postoperative neurologic grades at 3 to 7 days and 21 to 31 days after surgery and recovery times were lower in dogs with satisfactory versus unsatisfactory SCD. Moreover, dogs with unsatisfactory SCD were more likely to have an unsuccessful outcome or recurrence of neurologic signs. However, most (44/53 [83%]) dogs with satisfactory SCD had a preoperative neurologic grade of 1 or 2, which could have influenced the results. Previous reports6,36,37 of dogs with cervical and thoracolumbar disk herniation indicate an association between less severe preoperative neurologic grade and positive outcomes or faster improvement after surgery.
The results of the present study supported the use of postoperative MRI for dogs undergoing surgery for IVDH. In particular, postoperative MRI is suggested for patients with recurrence of clinical signs or slow improvement in neurologic condition after surgery given that unsatisfactory SCD may be responsible for residual clinical signs. In the present study, findings suggested an association between preoperative intramedullary hyperintensity on T2-weighted MRI images and postoperative neurologic grade at 3 to 7 days after surgery but no association with neurologic grade at 21 to 30 days after surgery, final outcome, or recovery time. Moreover, postoperative intramedullary hyperintensity was not associated with postoperative neurologic grade, final outcome, or recovery time. An association between intramedullary hyperintensity and unsuccessful outcome has been demonstrated previously, and this association has often been detected in dogs with a sudden onset of neurologic signs and dispersed disk extrusions.9,23,28,38
Differences between results of the present and previous studies may be related to differences in proportions of dogs with acute and subacute onset of neurologic signs (vs a chronic history of neurologic dysfunction), prevalences of dispersed (vs nondispersed) extruded disk material, and proportions of dogs with (vs without) intramedullary hyperintensity on T2-weighted MRI images. The use of low-field MRI in the present study may also have led to underestimation of changes in intraspinal signal.39–42 Interestingly, the number of dogs with intramedullary hyperintensity was lower in preoperative MRI images (12/68 [18%]) than in postoperative images (23/68 [34%]). These results may be explained by the removal of extruded material and better visibility of spinal cord parenchyma after SCD surgery.
Limitations of the present study included the small sample size and use of low-field MRI. Magnetic resonance imaging measurements, even though standardized and obtained in triplicate, were collected by 1 person only. Dogs with recurrence of clinical signs were immediately reevaluated neurologically to exclude sites of pain that were different from the previous surgical region or neurologic deficit, compatible with different neurolocalization. All of these dogs had a low severity of neurologic signs and improved with cage rest and medical management. Follow-up MRI could have yielded interesting data on factors inciting recurrence of neurologic signs but was deemed clinically unnecessary.
Overall, our results suggested that the surgeon's perception of adequate SCD may be less reliable than objective evaluation via postoperative MRI; postoperative imaging appeared to be particularly useful for dogs with a severe preoperative neurologic grade, severe preoperative SCC, and thoracolumbar IVDE. Postoperative MRI also allowed for identification of further intramedullary signal changes that may not have been visible on preoperative MRI images. Given the preliminary results reported here, further prospective investigation is warranted into the accuracy of the surgeon's intraoperative perception of SCD in dogs with cervical or thoracolumbar IVDH.
Acknowledgments
No third-party funding or support was received in connection with this study, data analysis and interpretation, or writing and publication of the manuscript. The authors declare that there were no conflicts of interest.
Presented in part as an oral presentation at the 31st Annual Symposium of the European Society of Veterinary Neurology, Copenhagen, September 2018.
The authors thank Drs. Liliana Carnevale and Gianluca Abbiati for their involvement in surgical procedures and Drs. Filippo Ferri and Federica Cantatore for providing technical support.
ABBREVIATIONS
| CSA | Cross-sectional area |
| ICC | Intraclass correlation coefficient |
| IVDE | Intervertebral disk extrusion |
| IVDH | Intervertebral disk herniation |
| SCC | Spinal cord compression |
| SCD | Spinal cord decompression |
| TE | Echo time |
| TR | Repetition time |
Footnotes
Hitachi Airis II, Hitachi Medical Systems, Milan, Italy.
Osirix MD, version 9.0.2, Pixmeo Sarl, Bernex, Switzerland.
Baytril Flavour, Bayer SpA, Milan, Italy.
Synulox, Pfizer Italia Srl, Milan, Italy.
Altadol, Formevet SpA, Milan, Italy.
Rimadyl, Pfizer Italia Srl, Milan, Italy.
Prednicortone, Dechra Veterinary Products, Turin, Italy.
SAS, version 9.4, SAS Institute Inc, Cary, NC.
References
1. Hakozaki T, Iwata M, Kanno N, et al. Cervical intervertebral disk herniation in chondrodystrophoid and nonchondrodystrophoid small-breed dogs: 187 cases (1993–2013). J Am Vet Med Assoc 2015;247:1408–1411.
2. Naudé SH, Lambrechts NE, Wagner WM, et al. Association of preoperative magnetic resonance imaging findings with surgical features in Dachshunds with thoracolumbar intervertebral disk extrusion. J Am Vet Med Assoc 2008;232:702–708.
3. Jeffery ND, Levine JM, Olby NJ, et al. Intervertebral disk degeneration in dogs: consequences, diagnosis, treatment, and future directions. J Vet Intern Med 2013;27:1318–1333.
4. Levine JM, Fingeroth JM. Historical and current nomenclature associated with intervertebral disc pathology. In: Fingeroth JM, Thomas WB, eds. Advances in intervertebral disc disease in dogs and cats. Ames, Iowa: John Wiley & Sons Inc, 2015;25–31.
5. Kazakos G, Polizopoulou ZS, Patsikas MN, et al. Duration and severity of clinical signs as prognostic indicators in 30 dogs with thoracolumbar disk disease after surgical decompression. J Vet Med A Physiol Pathol Clin Med 2005;52:147–152.
6. Ryan TM, Platt SR, Llabres-Diaz FJ, et al. Detection of spinal cord compression in dogs with cervical intervertebral disc disease by magnetic resonance imaging. Vet Rec 2008;163:11–15.
7. Cooper JJ, Young BD, Griffin JF, et al. Comparison between noncontrast computed tomography and magnetic resonance imaging for detection and characterization of thoracolumbar myelopathy caused by intervertebral disk herniation in dogs. Vet Radiol Ultrasound 2014;55:182–189.
8. Bos AS, Brisson BA, Nykamp SG, et al. Accuracy, intermethod agreement, and inter-reviewer agreement for use of magnetic resonance imaging and myelography in small-breed dogs with naturally occurring first-time intervertebral disk extrusion. J Am Vet Med Assoc 2012;240:969–977.
9. Besalti O, Pekcan Z, Sirin YS, et al. Magnetic resonance imaging findings in dogs with thoracolumbar intervertebral disk disease: 69 cases (1997–2005). J Am Vet Med Assoc 2006;228:902–908.
10. Jeffery ND, Harcourt-Brown TR, Barker AK, et al. Choices and decisions in decompressive surgery for thoracolumbar intervertebral disk herniation. Vet Clin North Am Small Anim Pract 2018;48:169–186.
11. Roach WJ, Thomas M, Weh JM, et al. Residual herniated disc material following hemilaminectomy in chondrodystrophic dogs with thoracolumbar intervertebral disc disease. Vet Comp Orthop Traumatol 2012;25:109–115.
12. Svensson G, Simonsson US, Danielsson F, et al. Residual spinal cord compression following hemilaminectomy and minihemilaminectomy in dogs: a prospective randomized study. Front Vet Sci 2017;4:42.
13. Levine JM, Levine GJ, Johnson SI, et al. Evaluation of the success of medical management for presumptive cervical intervertebral disk herniation in dogs. Vet Surg 2007;36:492–499.
14. Levine JM, Levine GJ, Johnson SI, et al. Evaluation of the success of medical management for presumptive thoracolumbar intervertebral disk herniation in dogs. Vet Surg 2007;36:482–491.
15. Hettlich BF, Kerwin SC, Levine JM. Early reherniation of disk material in eleven dogs with surgically treated thoracolumbar intervertebral disk extrusion. Vet Surg 2012;41:215–220.
16. Rossmeisl JH Jr, White C, Pancotto TE, et al. Acute adverse events associated with ventral slot decompression in 546 dogs with cervical intervertebral disc disease. Vet Surg 2013;42:795–806.
17. Forterre F, Gorgas D, Dickomeit M, et al. Incidence of spinal compressive lesions in chondrodystrophic dogs with abnormal recovery after hemilaminectomy for treatment of thoracolumbar disc disease: a prospective magnetic resonance imaging study. Vet Surg 2010;39:165–172.
18. Böttcher P, Böttcher IC, Ludewing E. Effect of ventral slot procedure on spinal cord compression in dogs with single static intervertebral disc disease: preliminary findings while evaluating a semiquantitative computed tomographic myelographic score of spinal cord compression. Vet Surg 2013;42:383–391.
19. Huska JL, Gaitero L, Brisson BA, et al. Presence of residual material following mini-hemilaminectomy in dogs with thoracolumbar intervertebral disc extrusion. Can Vet J 2014;55:975–980.
20. Salger F, Ziegler L, Böttcher IC. Neurologic outcome after thoracolumbar partial lateral corpectomy for intervertebral disc disease in 72 dogs. Vet Surg 2014;43:581–588.
21. Laflamme DRPC. Development and validation of a body condition score system for dogs. Canine Pract 1997;22(1):10–15.
22. Martínez S, Fajardo R, Valdés J, et al. Histopathologic study of long-bone growth plates confirms the basset hound as an osteochondrodysplastic breed. Can J Vet Res 2007;71:66–69.
23. Levine JM, Fosgate GT, Chen AV, et al. Magnetic resonance imaging in dogs with neurologic impairment due to acute thoracic and lumbar intervertebral disk herniation. J Vet Intern Med 2009;23:1220–1226.
24. Bergknut N, Auriemma E, Wijsman S, et al. Evaluation of intervertebral disk degeneration in chondrodystrophic and nonchondrodystrophic dogs by use of Pfirrmann grading of images obtained with low-field magnetic resonance imaging. Am J Vet Res 2011;72:893–898.
25. Kranenburg HJ, Grinwis GC, Bergknut N, et al. Intervertebral disc disease in dogs—part 2: comparison of clinical, magnetic resonance imaging, and histological findings in 74 surgically treated dogs. Vet J 2013;195:164–171.
26. Smolders LA, Bergknut N, Grinwis GC, et al. Intervertebral disc degeneration in the dog. Part 2: chondrodystrophic and non-chondrodystrophic breeds. Vet J 2013;195:292–299.
27. Pfirrmann CW, Metzdorf A, Zanetti M, et al. Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine (Phila Pa 1976) 2001;26:1873–1878.
28. Besalti O, Ozak A, Pekcan Z, et al. The role of extruded disk material in thoracolumbar intervertebral disk disease: a retrospective study in 40 dogs. Can Vet J 2005;46:814–820.
29. Beltran E, Dennis R, Doyle V, et al. Clinical and magnetic resonance imaging features of canine compressive cervical myelopathy with suspected hydrated nucleus pulposus extrusion. J Small Anim Pract 2012;53:101–107.
30. Sharp NJH, Wheeler SJ. Cervical disc disease. In: Sharp NJH, Wheeler SJ, eds. Small spinal disorders: diagnosis and surgery. Philadelphia: Elsevier Ltd, 2005;92–120.
31. Sharp NJH, Wheeler SJ. Thoracolumbar disc disease. In: Sharp NJH, Wheeler SJ, eds. Small animal spinal disorders: diagnosis and surgery. Philadelphia: Elsevier Ltd, 2005;121–159.
32. Aikawa T, Fujita H, Shibata M, et al. Recurrent thoracolumbar intervertebral disc extrusion after hemilaminectomy and concomitant prophylactic fenestration in 662 chondrodystrophic dogs. Vet Surg 2012;41:381–390.
33. Forterre F, Konar M, Spreng D, et al. Influence of intervertebral disc fenestration at the herniation site in association with hemilaminectomy on recurrence in chondrodystrophic dogs with thoracolumbar disc disease: a prospective MRI study. Vet Surg 2008;37:399–405.
34. Penning V, Platt SR, Dennis R, et al. Association of spinal cord compression seen on magnetic resonance imaging with clinical outcome in 67 dogs with thoracolumbar intervertebral disc extrusion. J Small Anim Pract 2006;47:644–650.
35. Dhupa S, Glickman N, Waters DJ. Reoperative neurosurgery in dogs with thoracolumbar disc disease. Vet Surg 1999;28:421–428.
36. Cherrone KL, Dewey CW, Coates JR, et al. A retrospective comparison of cervical intervertebral disk disease in non-chondrodystrophic large dogs versus small dogs. J Am Anim Hosp Assoc 2004;40:316–320.
37. Ingram EA, Kale DC, Balfour RJ. Hemilaminectomy for thoracolumbar Hansen Type I intervertebral disk disease in ambulatory dogs with or without neurologic deficits: 39 cases (2008–2010). Vet Surg 2013;42:924–931.
38. De Risio L, Adams V, Dennis R, et al. Association of clinical and magnetic resonance imaging findings with outcome in dogs with presumptive acute noncompressive nucleus pulposus extrusion: 42 cases (2000–2007). J Am Vet Med Assoc 2009;234:495–504.
39. Henke D, Gorgas D, Flegel T, et al. Magnetic resonance imaging findings in dogs with traumatic intervertebral disk extrusion with or without spinal cord compression: 31 cases (2006–2010). J Am Vet Med Assoc 2013;242:217–222.
40. Maubon AJ, Ferru JM, Berger V, et al. Effect of field strength on MR images: comparison of the same subject at 0.5, 1.0, and 1.5 T. Radiographics 1999;19:1057–1067.
41. Magee T, Shapiro M, Williams D. Comparison of high-field-strength versus low-field-strength MRI of the shoulder. AJR Am J Roentgenol 2003;181:1211–1215.
42. Przeworski A, Adamiak Z, Głodek J. Comparison of high-field and low-field magnetic resonance imaging of stifle joint disorders in dogs. Pol J Vet Sci 2016;19:663–670.
