A 3-year-old 10-kg (22-lb) spayed female French Bulldog (dog 1) was presented because of a sudden onset of signs of pain. The dog had been confined in a crate during the day but, when released, it cried out when jumping onto the owner. On physical and neurologic examination, the dog was considered normal except for signs of pain elicited on palpation of the region over the caudal lumbar vertebrae. Treatment of the dog included administration of gabapentin (10 mg/kg [4.5 mg/lb], PO, q 8 to 12 h), tramadol (5 mg/kg [2.3 mg/lb], PO, q 8 to 12 h), and prednisone (0.5 mg/kg [0.23 mg/lb], PO, q 12 h) and strict exercise restriction. The next day, the dog had difficulty walking. On reevaluation, the dog was ambulatory but paraparetic. The pelvic limbs had delayed postural reactions, reduced patellar and withdrawal reflexes, and flaccid muscle tone. The perineal reflex and external anal sphincter tone were diminished. The remainder of the examination findings were considered normal. Signs were consistent with a lesion affecting the L4 through S3 spinal cord segments or roots, L4 through S3 spinal nerves, or the nerves of the pelvic limbs (ie, femoral and sciatic nerves) and the pudendal nerve. Differential diagnoses included IVDH, myelitis (infectious or immune mediated), and neoplasia. Results of a CBC were within reference intervals. Serum biochemical analysis revealed high potassium concentration (5.4 mmol/L; reference interval, 3.7 to 4.7 mmol/L) and high alkaline phosphatase activity (444 U/L; reference interval, 11 to 133 U/L).
A 6-year-old 30-kg (66-lb) neutered male mixed-breed dog (dog 2) was presented because of a sudden onset of paraparesis. Signs began with right pelvic limb lameness and progressed in 24 hours to ambulatory paraparesis. The dog had a 2-year history of diabetes mellitus, which was controlled with insulina (18 U, SC, q 12 h). Physical examination findings were considered normal. On neurologic examination, the dog was ambulatory with a general proprioceptive ataxia and upper motor neuron paraparesis. The pelvic limbs had delayed postural reactions, increased patellar and normal withdrawal reflexes, and spastic muscle tone. Signs of pain were elicited on palpation of the region over the cranial lumbar vertebrae. The remainder of the neurologic examination findings were considered normal. The clinical signs were consistent with a lesion involving the T3 to L3 spinal cord segments. Differential diagnoses were the same as those for dog 1. A CBC revealed no abnormalities. Serum biochemical analysis revealed high activities of alkaline phosphatase (174 U/L; reference interval, 20 to 150 U/L) and alanine transferase (133 U/L; reference interval, 10 to 118 U/L) and high glucose concentration (485 mg/dL; reference interval, 60 to 110 mg/dL).
A 6-year-old 11.2-kg (24.6-lb) neutered male mixed-breed dog (dog 3) was presented because of a sudden onset of nonambulatory paraparesis. Two days earlier, the dog was reluctant to walk and had a kyphotic posture. The dog was treated by the referring veterinarian with administration of methocarbamol (22.3 mg/kg [10.1 mg/lb], PO, q 12 h), gabapentin (8.9 mg/kg [4.0 mg/lb], PO, q 12 h), and carprofen (2.2 mg/kg [1.0 mg/lb], PO, q 12 h) and exercise restriction. Despite treatment, the dog's condition progressed to nonambulatory paraparesis. Physical examination findings for the dog were considered normal. On neurologic examination, the dog had nonambulatory paraparesis. The pelvic limbs had delayed postural reactions, normal patellar and withdrawal reflexes, and spastic muscle tone. Signs of pain were elicited on palpation of the region over the cranial lumbar vertebrae. The remainder of the neurologic examination findings were considered normal. The clinical signs were consistent with a lesion involving the T3 to L3 spinal cord segments. The differential diagnoses were the same as those for dog 1. Results of a CBC were within reference intervals. Serum biochemical analysis revealed high concentrations of total protein (8.1 g/dL; reference interval, 5.0 to 7.3/dL) and albumin (4.8 g/dL; reference interval, 2.2 to 3.9 g/dL).
Prior to induction of anesthesia for MRI and surgery, all dogs had normal hydration status. All dogs received midazolam (0.2 mg/kg [0.9 mg/lb], IV) and hydromorphone (0.1 mg/kg [0.045 mg/lb], IV) or methadone (0.5 mg/kg, IV). Anesthesia was induced with propofol (2.0 to 4.0 mg/kg [0.9 to 1.8 mg/lb], IV) or ketamine hydrochloride (5 mg/kg, IV) and midazolam (0.25 mg/kg [0.11], IV) and was maintained via inhalation of isoflurane and oxygen. Isoflurane vaporizer settings ranged from 1% to 2%. During most of the period of anesthesia, each dog was mechanically ventilated. Lactated Ringer solution was administered at 5 mL/kg/h. The following variables were recorded every 5 minutes: heart rate (recorded as beats/min), respiratory rate (recorded as breaths/min), body temperature (recorded via esophageal thermometer), arterial blood pressure, ETco2, oxygen saturation (as measured by pulse oximetry), oxygen flow rate, and isoflurane vaporizer setting. Physiologic variables were obtained by use of multivariable patient monitoring systems.b,c Arterial blood pressure was directly measured via catheterization of the dorsal pedal artery, and SAP, MAP, and DAP were recorded. Indirect blood pressure measurements were obtained with Doppler ultrasonography and a sphygmomanometer and a blood pressure cuff measuring approximately 40% of the circumference of the dog's antebrachium. With Doppler ultrasonography, only SAP measurements were obtained.
All dogs underwent MRI of the thoracolumbar portion of the vertebral column from the T1 vertebra to the cranial aspect of the sacrum with a 3.0-T MRI unit.d Multiplanar T2-weighted and T1-weighted images were acquired. Additional T1-weighted images were acquired after IV administration of gadopentetate dimegluminee (0.1 mmol/kg). Postcontrast T1-weighted images were acquired with the Dixon method of fat suppression. Imaging settings for T2-weighted sequences were as follows: echo time = 9.4 to 12 milliseconds, repetition time = 2,610 to 6,990 milliseconds, echo train = 17 to 19, slice thickness = 2 to 3 mm, interslice gap = 0 to 0.3 mm, and square field of view = 100 to 280 × 100 to 280 mm; the matrix ranged from 256 to 640 (frequency encoding) by 218 to 384 (phase encoding). Imaging settings for T1-weighted sequences were as follows: echo time = 9.4 to 15 milliseconds, repetition time = 400 to 731 milliseconds, echo train = 3, slice thickness = 2 to 3 mm, interslice gap = 0 to 0.3 mm, and square field of view = 120 to 300 × 120 to 300 mm; the matrix ranged from 320 to 704 (frequency encoding) by 224 to 369 (phase encoding).
In all dogs, there was extradural material in the right ventral portion of the epidural space at the L3-4 intervertebral space that resulted in spinal cord compression (Figure 1). Subjectively, the degree of compression was considered mild for dog 1 and severe for dogs 2 and 3. On T2-weighted and T1-weighted images, the material was heterogeneously iso- to hypointense. In dog 3, the material also had areas of hyperintensity on T2-weighted and T1-weighted images, which were consistent with accumulations of blood breakdown products. Following administration of contrast medium, no additional pathological changes were evident in any dog.
Preoperative sagittal (A, C, E) and corresponding preoperative transverse (B, D, F) T2-weighted MRI images at approximately the level of the L3-4 intervertebral space of dogs 1 (A and B), 2 (C and D), and 3 (E and F) that were initially examined because of a sudden onset of signs of pain (dog 1) or paraparesis (dogs 2 and 3). In all dogs, there is extradural material (arrows) causing compression of the spinal cord.
Citation: Journal of the American Veterinary Medical Association 257, 4; 10.2460/javma.257.4.397
All 3 dogs underwent a standard right-sided hemilaminectomy at the L3-4 vertebral articulation. The right side was chosen on the basis of the location of the extradural material observed on MRI images. Briefly, following dorsal midline skin incision, the subcutaneous tissues were sharply dissected and the right epaxial muscles were elevated off of the spinous process, laminae, and pedicles from the L2 through L4 vertebrae. The articular processes at the L3-4 articulation were removed with a bone rongeur, and a high-speed pneumatic drill was used to create the hemilaminectomy centered over the L3-4 intervertebral space. In each dog, there was compression of the spinal cord by epidural hemorrhage admixed with material that had the visual characteristics of degenerative nucleus pulposus. The compressive extradural material was removed without need for excessive manipulation of the spinal cord by use of a right-angle nerve hook and Billeau ear loop. Following removal of the extradural material, the meninges and spinal cord appeared grossly normal. None of the dogs had excessive hemorrhage during surgery. The surgical site was closed routinely without the use of a fat graft or placement of a hemostatic sponge. Complications during surgery were not encountered in any dog.
The median duration of anesthesia, MRI, and surgery was 5 hours (range, 4 hours 13 minutes to 5 hours 13 minutes), 53 minutes (range, 40 minutes to 1 hour), and 2 hours 10 minutes (range, 2 hours 7 minutes to 2 hours 20 minutes), respectively. In all dogs, direct blood pressure measurements were obtained after induction of anesthesia and during acquisition of the MRI images (Figure 2). Blood pressure was directly measured for 75 minutes in all dogs. During the remainder of the entire period of anesthesia, blood pressure was measured indirectly. When blood pressure was measured invasively, the median SAP, MAP, and DAP were 100 mm Hg (range, 75 to 120 mm Hg), 77 mm Hg (range, 60 to 110 mm Hg), and 57 mm Hg (range, 40 to 100 mm Hg), respectively. The lowest MAP of 60 mm Hg was recorded only once in dogs 1 and 2. The lowest SAP of 75 mm Hg was only recorded once in dog 3. In dog 2, an SAP of 80 mm Hg was recorded twice, but those recordings were not consecutive.
Plots of physiologic variables monitored during anesthesia for MRI and surgery for dogs 1 (A), 2 (B), and 3 (C). Variables that were monitored over the approximately 5-hour period included heart rate (purple line), SAP (direct measurement, orange line; indirect measurement, blue line), MAP (direct measurement, gray line), DAP (direct measurement, yellow line), ETco2 (green line), respiration rate (red line), and esophageal temperature (brown line). Administration of atropine (arrows) and IV fluid boluses (arrowheads) are denoted. Brackets denote the transition from the MRI suite to the operating room.
Citation: Journal of the American Veterinary Medical Association 257, 4; 10.2460/javma.257.4.397
During the entire period of anesthesia, the median heart rate among the 3 dogs was 65 beats/min (range, 35 to 175 beats/min). In dog 1, two doses of atropine sulfate (0.01 mg/kg [0.005 mg/lb], IV) were administered 5 minutes apart when the heart rate was 35 beats/min (Figure 2). Contemporaneously, the MAP was 60 mm Hg measured directly for a single recording (SAP was not recorded). Atropine administration increased the heart rate, which then remained > 65 beats/min for the remainder of the period of anesthesia. In dog 3, one dose of atropine (0.01 mg/kg, IV) was administered when the heart rate was 45 beats/min. Contemporaneously, the directly measured SAP and MAP were 75 and 65 mm Hg, respectively. Following the administration of atropine, the heart rate remained > 65 beats/min, and the directly measured SAP and MAP were > 80 mm Hg and > 60 mm Hg, respectively, for the remainder of the period of anesthesia. Dog 2 did not receive atropine. Dog 1 received 2 IV fluid boluses (5 mL/kg) 30 minutes apart when the heart rate was > 60 beats/min and the indirectly measured SAP was > 90 mm Hg. The reason for administration of the IV fluid boluses was not recorded. Dog 3 received a single IV fluid bolus (5 mL/kg) 30 minutes after induction of anesthesia when the indirectly measured SAP was 120 mm Hg and the heart rate was 55 beats/min. The reason cited for the fluid bolus administration was the dog's high total protein and albumin concentrations.
All dogs were mechanically ventilated during most of the period of anesthesia; therefore, respiratory rate varied little (Figure 2). Dogs 1, 2, and 3 were ventilated at 10 to 15 breaths/min, 10 to 12 breaths/min, and 8 to 12 breaths/min, respectively. The median ETco2 was 40 mm Hg (range, 24 to 69 mm Hg). For dog 1, an ETco2 of 69 mm Hg was recorded when the dog was moved into the operating room. Contemporaneously, the heart rate was 65 beats/min, and the indirect SAP measurement was 90 mm Hg. In 2 recording periods, the ETco2 for dog 1 was < 45 mm Hg. For dog 3, mechanical ventilation was started when the ETco2 was 57 mm Hg while the dog was breathing spontaneously. With mechanical ventilation, the ETco2 for dog 3 was 24 mm Hg and remained < 30 mm Hg for 7 recording periods, during which time adjustments to the rate and volume of ventilation were made to establish ETco2 > 30 mm Hg. During that adjustment time, both the direct SAP and MAP measurements were > 80 mm Hg, and heart rate was between 50 and 60 beats/min. For dog 2, the ETco2 was maintained between 35 and 44 mm Hg. Oxygen saturation as measured by pulse oximetry, oxygen flow rate, and isoflurane concentration were > 97%, 1.0 to 2.0 L/min, and 1.0% to 1.5%, respectively. The median body temperature among the 3 dogs was 37.0°C (98.6°F; range, 36.2°C to 38.1°C [97.2°F to 100.7°F]).
During surgery, all dogs were administered fentanyl (5 to 10 μg/kg/h, IV) as a continuous rate infusion with adjustments made on the basis of anesthetic depth and reaction to the surgical procedure. Dog 2 received ketamine (1.2 mg/kg/h [0.55 mg/lb/h], IV) and lidocaine (3 mg/kg/h [1.36 mg/lb/h], IV) as a continuous rate infusion during surgery.
All dogs recovered from anesthesia without excessive signs of pain, anxiety, or movements. All dogs were maintained on IV fluids and fentanyl (2 to 3 μg/kg/h, IV) as a continuous rate infusion with adjustments made on the basis of subjective assessment of their pain control. Neurologic examinations were performed between 12 and 16 hours following surgery for all dogs. All dogs had flaccid paraplegia; absent patellar, withdrawal, and perineal reflexes; flaccid external anal sphincter and tail tone; and absent nociception in the pelvic limbs. The remainder of the neurologic examination findings were considered normal. The clinical signs were consistent with a lesion affecting the L4 through caudal spinal cord segments, the L4 through caudal spinal roots or spinal nerves, or the femoral, sciatic, pelvic, pudendal, and caudal nerves.
Owing to decline of neurologic function, repeated MRI (with the same anesthetic protocol and MRI unit) was performed for dogs 1, 2, and 3 at 17.7, 14.5, and 24 hours after surgery, respectively (Figure 3). The repeated MRI was performed to assess for spinal cord compression from residual or newly extruded IVD material. For all dogs, T2-weighted images were acquired; other pulse sequences were acquired at the discretion of the attending clinicians. The MRI findings were similar in all dogs. The subcutaneous tissue at the incision site was disrupted. The right epaxial muscles were swollen and had illdefined hyperintensity on the T2-weighted images with variable contrast enhancement along the hemilaminectomy site, which was consistent with surgical intervention. Observed focal signal voids were consistent with hemorrhage, gas, osseous fragments, or artifacts from the metal burr.1
Postoperative sagittal (A, C, E) and corresponding postoperative transverse (B, D, F) T2-weighted MRI images at approximately the level of the L3-4 intervertebral space of dogs 1 (A and B), 2 (C and D), and 3 (E and F). All 3 dogs underwent a standard right-sided hemilaminectomy at the L3-4 vertebral articulation. In the sagittal images (A, C, E), hyperintensity of the spinal cord (arrowhead) cranial and caudal to the L3-4 intervertebral space (arrow) is observed. On the transverse T2-weighted images (B, D, F), the area of the spinal cord gray matter is hyperintense (arrows). The epaxial muscles are disrupted with edema, and there is focal perivertebral hemorrhage and fluid and gas accumulation at the site of the right-sided hemilaminectomy.
Citation: Journal of the American Veterinary Medical Association 257, 4; 10.2460/javma.257.4.397
In all dogs, the lumbar portion of the spinal cord was hyperintense on the T2-weighted images and subjectively swollen, compared with the previous MRI findings. The region of hyperintensity extended from T11 (dog 1), L1 (dog 2), and L3 (dog 3) to the termination of the spinal cord. On transverse T2-weighted images of the L2 or L3 vertebra through the L5 vertebra, the hyperintensity was associated with the gray matter of the spinal cord. Subjectively, the spinal cord was considered decompressed in dogs 2 and 3. However, in dog 1, extradural material at the L3-4 intervertebral space was present and had caused a greater degree of spinal cord compression, compared with the initial MRI findings. After the repeated MRI, dog 1 underwent an exploratory surgery during which hemorrhage admixed with material that had the gross appearance of degenerative nucleus pulposus was removed. The findings were interpreted as continued extrusion of degenerative IVD material. In dogs 2 and 3, the spinal cord at the L3-4 intervertebral space appeared decompressed with minimal residual material in the epidural space.
Neurologic improvement was not observed during hospitalization in any dog. Dogs 1 and 2 were discharged from the hospital 19 and 25 days after surgery. Three days after surgery, dog 3 was euthanized by IV administration of an overdose of pentobarbital while in the hospital. Dogs 1 and 2 were euthanized by IV administration of an overdose of pentobarbital at 91 and 34 days, respectively, owing to a lack of neurologic improvement. All dogs underwent postmortem gross and histologic evaluation of the spinal cord.
For all dogs, the spinal cord was removed from the vertebral column within 24 hours after euthanasia. The spinal cord was preserved in neutral-buffered 10% formalin. Once preserved, representative transverse sections of the lumbar portion of the spinal cord were obtained, routinely processed, and stained with H&E stain.
Grossly, the meninges and spinal cord appeared normal in all dogs. On cut section, the region of the gray matter of the lumbar portion of the spinal cord in dogs 1 and 2 was light brown. The discolored region of the gray matter extended from the L4 (dog 1) and L2 (dog 2) spinal cord segments to the termination of the spinal cord. In dog 3, cut sections were grossly normal. Microscopically, the cranial segments of the affected spinal cord in dogs 1 and 2 had focal areas of rarefied gray matter; more caudally, the extent of the affected area of the gray matter increased such that the entirety of the gray matter at the L6 (dog 1) and L4 (dog 2) spinal cord segments was rarefied. The affected gray matter in dogs 1 and 2 was nearly devoid of neurons. A few neurons were present at the gray matter–white matter junction in the dorsal aspect of the dorsal horn in dog 1. The area previously occupied by gray matter neurons contained lipid-laden macrophages (gitter cells) and mononuclear cells that were presumed to be microglia and lymphocytes. In dog 2, there was degeneration of the white matter immediately adjacent to the affected gray matter surrounded by a peripheral rim of less affected white matter. In dogs 1 and 2, there was degeneration of the ventral spinal nerve rootlets and roots (Figure 4).
Photomicrographs of transverse sections of the spinal cord obtained at approximately the level of the L3-4 intervertebral space in dogs 1 (A), 2 (B), and 3 (C) following euthanasia. A—In dog 1, there is severe rarefaction of the entirety of the gray matter. Notice the relative preservation of white matter. Compared with the dorsal nerve rootlets (open arrow), degeneration of the ventral spinal nerve rootlets (black arrow) secondary to a loss of ventral horn neurons is present. The box denotes a region illustrated in the inset. H&E stain; bar = 500 μm. Inset—The ventral gray matter neuropil has been completely replaced with macrophages filled with lipid (gitter cells) and lymphocytes. H&E stain; bar = 20 μm. B—In dog 2, the entirety of the gray matter has similar rarefaction, which extends into the white matter immediately surrounding the gray matter. The box denotes a region illustrated in the inset. H&E stain; bar = 500 μm. Inset—In the area of the ventrally located gray matter, only macrophages filled with lipid (gitter cells) are present. H&E stain; bar = 20 μm. C—In dog 3, there is an increase in cellularity and slight vacuolation of the spinal cord gray matter (open arrow). The box denotes a region illustrated in the insets. H&E stain; bar = 500 μm. Top inset—Two neurons in the ventral portion of the gray matter have eosinophilic cytoplasm with degeneration of the nucleus (black arrows). Another neuron (arrowhead) has chromatolysis. A capillary with reactive endothelium also is present (open arrow). H&E stain; bar = 20 μm. Bottom inset—Other neurons (black arrows) in the ventral portion of the spinal cord gray matter have a more classic appearance of ischemic cell change. There are shrunken, angular neurons that lack Nissl substance and have hypereosinophilic cytoplasm and a condensed, hyperchromatic nucleus. H&E stain; bar = 20 μm.
Citation: Journal of the American Veterinary Medical Association 257, 4; 10.2460/javma.257.4.397
In dog 3, the changes in the gray matter were less well developed. From L3 through to the termination of the spinal cord, the neuropil had focal areas of vacuolation that were infiltrated by neutrophils and mononuclear cells. Hypertrophied blood vessels lined by reactive endothelium were present throughout the gray matter. Neurons frequently had abnormal Nissl substance and small hyperchromatic nuclei. In the L4 spinal cord segment, the neurons in the ventral horn were angular with hypereosinophilic cytoplasm and pyknotic nuclei (acidophilic neuronal necrosis; Figure 4). In all dogs, evidence of spinal cord hemorrhage was absent.
Discussion
The 3 dogs of the present report had a dramatic decline in neurologic function following routine anesthesia for MRI of the vertebral column and decompressive surgery of an IVDH. Postoperative MRI of the vertebral column revealed hyperintensity involving the gray matter of the lumbar intumescence in T2-weighted images. The location and extent of the hyperintensity correlated with the postmortem microscopic changes, which consisted of poliomyelomalacia that severely affected the gray matter of the lumbar intumescence with less extensive degeneration of the adjacent white matter. Degeneration of the spinal cord gray matter with sparing of the peripheral white matter was consistent with an ischemic spinal cord injury, as seen in rabbits and cats that develop diffuse poliomyelomalacia following experimental occlusion of the aorta.2,3 Similar clinical and microscopic findings have been described for horses and a calf following anesthesia4,5 and for humans who have undergone surgical manipulation or traumatic laceration of the aorta.6 To the authors' knowledge, such extensive poliomyelomalacia associated with anesthesia and surgery for IVDH in dogs has not been previously described.
Similar microscopic changes in dogs with chronic (> 2 months' duration) neurologic dysfunction secondary to IVDH have been reported.7 However, in those cases, the clinical course was not described, and the gray matter degeneration was centered at the site of spinal cord compression where there was marked deformation of the cross-sectional shape of the spinal cord. In contradistinction, the gray matter changes in the dogs of the present report were acute. Also, the gray matter degeneration extended along several spinal cord segments distant from the site of compression by the IVDH, and there was little deformation of the shape of the spinal cord, supporting an underlying ischemic etiopathogenesis.
In dogs, a common ischemic myelopathy is FCEM.8 In most instances, cartilaginous emboli are observed in the meningeal vasculature or the intrinsic vasculature of the spinal cord.9 Unlike the topography of the ischemic lesion observed in the dogs of the present report, FCEM typically affects both gray and white matter and is often confined to the ventral or dorsal aspect of the spinal cord; however, FCEM uncommonly affects the gray matter solely.9 Cartilaginous emboli were not identified in any of the dogs of the present report. Additionally, although degenerative changes can be observed in the IVDs in close proximity to the affected region of the spinal cord, FCEM does not occur concurrently with acute extrusion of IVD material. Consequently, it is unlikely that the findings in the present cases represented FCEM.
Typical gross and microscopic changes in the spinal cord of dogs secondary to IVDH are categorized as compression (a distortion of the spinal cord), malacia (nervous tissue necrosis with the development of cystic spaces), or diffuse degenerative change in the white matter.10 Degenerative changes in the white matter consist of rarefaction as a consequence of dilation of myelin sheaths, swollen axons, and axonal loss, all of which affect funiculi principally at the level of the IVDH.11–13 White matter degeneration may extend contiguously or as separate foci from the primary site of IVDH.10,11 Chronically, degenerative white matter may develop cyst-like cavitations.7,14 When present, neuronal loss involving the dorsal, intermediate, or ventral regions of gray matter is frequently accompanied by degeneration of the white matter.10,12,13 In some instances, pan-necrosis of the gray matter and white matter may develop.11
The high metabolic demand and vascularity of the gray matter makes it particularly vulnerable to ischemia.5,15 Possible causes of the ischemic poliomyelomalacia in the dogs of the present report included disruption of arterial blood supply to the spinal cord, impaired venous drainage, hypoxia secondary to hypotension, or decreased cardiac output. Because all dogs had a normal hydration status at the time of anesthesia and surgery, none of the medications administered to the dogs or the presence of diabetes mellitus in dog 3 was considered to have contributed to the observed ischemia.
In horses with postanesthetic poliomyelomalacia, disruption of the arterial blood supply may be a result of compression of the aorta by abdominal viscera when the horses were positioned in dorsal recumbency.4 The dogs of the present report were positioned in dorsal recumbency when undergoing MRI but thereafter were placed in sternal recumbency. While each dog was in dorsal recumbency, arterial blood pressure measured directly from the dorsal pedal artery of a pelvic limb suggested unimpeded aortic blood flow. In humans, compression of a segmental spinal artery or the ventral spinal artery can result in ischemic myelopathy that principally affects gray matter.16 To create neurologic dysfunction related to spinal cord ischemia in dogs, multiple segmental spinal arteries need be occluded.17 The extensive poliomyelomalacia in the dogs of the present report suggested that the ischemia had developed over a large region of the spinal cord rather than from occlusion of a single segmental vessel. In cats, traumatic injuries not associated with vertebral fractures, luxations, or spinal cord compression can result in ischemic myelopathy that affects the gray matter of the lumbar, sacral, and caudal spinal cord segments and is strikingly similar to the gross and microscopic findings in the dogs of the present report.18 Traumatic injury in cats, perhaps as a result of hypaxial muscle swelling that compresses the abdominal aorta or multiple lumbar arteries and compromises blood flow to the spinal cord, is thought to underlie development of the ischemic myelopathy.18
Alternatively, ischemia may have been related to venous outflow obstruction because spinal cord perfusion is determined by arterial blood inflow balanced against outflow via venous drainage and subarachnoid pressure.19 Obstruction of venous outflow from the spinal cord has been posited to underlie postanesthetic myelopathy in horses.20 Similar to the dogs of the present report, lesions in horses with postanesthetic myelopathy are primarily restricted to the spinal cord gray matter.4 Although the dogs were routinely positioned for MRI and surgery, it is possible that venous obstruction may have occurred in these dogs as a result of compression of the vena cava. Likewise, obstruction of the internal vertebral venous plexus may have occurred during or after surgery. None of the dogs had excessive venous hemorrhage during surgery or had a subcutaneous fat graft or a hemostatic gelatin sponge placed over the hemilaminectomy window that could account for obstruction of the internal vertebral venous plexus.
In the dogs of the present report, ischemia may have developed during anesthesia secondary to systemic hypotension or decreased cardiac output.21,22 Adverse events (including a decline in neurologic function) have been associated with hypotension in dogs undergoing decompressive surgery for IVDH, as defined by direct measurements of MAP ≤ 60 mm Hg or indirect measurements of SAP ≤ 80 mm Hg on ≥ 2 consecutive readings 5 minutes apart.23 Factors such as duration of anesthesia, bradycardia, mean body temperature, and mean ETco2 also may affect the outcome in dogs following decompressive surgery for IVDH.24,25 Arterial blood pressure does affect spinal cord hemodynamics.26 Mean SAP is positively correlated with spinal cord blood flow and perfusion pressure.27 For the dogs of the present report, it was difficult to isolate a time point when ischemia may have occurred on the basis of the recorded physiologic variables.
Spinal cord perfusion also is negatively impacted by increases in intraparenchymal and subarachnoid pressure, which may be present with spinal cord compression.28 Increased intraparenchymal or subarachnoid pressure combined with changes in other physiologic variables29 may have caused spinal cord ischemia in the dogs of the present report. Ultimately, multiple factors rather than a single influence likely contributed to development of poliomyelomalacia.
In the cases described in the present report, the decision to perform repeated MRI was based on the sudden and severe decline in neurologic function of the dogs, wherein all dogs were paraplegic and had absent nociception in the pelvic limbs following surgery. Although a slight postoperative decline in neurologic function may occur, the lack of nociception in the dogs of the present report was interpreted as too dramatic, compared with any expected change. Potential causes for postoperative decline in neurologic function include spinal cord compression secondary to continued extrusion of IVD material or extradural hemorrhage or hematoma formation. Although repeated anesthesia for MRI may pose a potential risk of further spinal cord injury, the possible benefit of identifying compression that could be relieved was felt to outweigh the potential negative impact anesthesia may have had on the spinal cord of the dogs in the present report. If spinal cord compression was identified after surgery, as initially observed in dog 1, surgical decompression of the spinal cord would have been considered. In dogs with IVDH that lack nociception, surgical decompression appears to offer improved prognosis over conservative treatment.30 In the setting of early postoperative decline in neurologic function, monitoring patients for improved neurologic function over several days may offer an alternative and more conservative approach. Although there is sparse literature-based support for early reoperation in dogs with continued IVD extrusion, reoperation may result in improved functional recovery.31–33 Clinicians should carefully weigh the potential benefits of early repeated imaging and reoperative decompressive surgery against the potential for causing further injury to the spinal cord from a second anesthesia event and decompressive surgery.
The cases described in the present report have highlighted that poliomyelomalacia secondary to ischemia may develop in dogs that experience a dramatic decline in neurologic function after surgery. Early postoperative MRI revealed hyperintensity of the spinal cord gray matter on T2-weighted images that appeared to indicate the development of ischemia. Dogs 1 and 2 had 91 and 34 days of recovery, respectively, and neither dog showed any improvement in neurologic examination findings over that time. Had a similar loss of gray matter with preservation of the white matter occurred within the T3 to L3 spinal cord segments, recovery of ambulation may have been possible.34,35 Typically, if dogs with thoracolumbar IVDH that lack nociception regain the ability to walk, they do so by 4 weeks after surgery.35,36 It is anticipated that ischemia affecting the gray matter of the intumescence would be associated with a poor prognosis, as evident in 2 of the dogs of the present report. If substantiated, an association of hyperintensity of the gray matter in T2-weighted images and the possibility that such changes may represent ischemia in affected dogs may provide clinicians with data to help advise owners of the potential for long-term or permanent neurologic deficits or prolonged hospitalization of their pet and added financial costs.
The microscopic findings for dog 3 differed from those for dogs 1 and 2 likely as a result of the time of euthanasia relative to the IVDH and decompression surgery. In dog 3, neurons in the spinal cord gray matter had ischemic changes rather than neuronal loss and replacement by macrophages as seen in dogs 1 and 2. The ischemic neurons in dog 3 supported the contention that ischemia was the underlying cause of poliomyelomalacia in the cases described in the present report. It was considered likely that, given more time, changes in the affected spinal cord gray matter in dog 3 would have progressed and that the appearance of the gray matter would have become similar to that in the other 2 dogs. In horses with postanesthetic ischemic myelopathy, the interval from cessation of anesthesia to euthanasia affects the severity of histopathologic lesions.4 Moreover, a time course for the development of the microscopic changes may be inferred by comparison of findings for dog 3 versus dog 2 and dog 2 versus dog 1, wherein the severity of changes worsened with increasing time between surgery and euthanasia.
The main limitations of the present report were its retrospective nature, small case number, variable periods of follow-up, lack of documented evaluation of urinary bladder function or coagulation testing, and the fact that physiologic variables were not measured by a single method throughout all anesthetic episodes. Arterial blood pressure variables are more accurately represented by direct monitoring methods than by indirect monitoring methods. For the dogs of the present report, direct measurements of arterial blood pressure appeared somewhat erratic, compared with the indirect measurements. This may have reflected a difference in positioning (ie, dorsal recumbency for MRI and sternal recumbency for surgery) or the depth of anesthesia, given that direct monitoring of arterial blood pressure was used at the onset of anesthesia when the dogs may have been in a light anesthetic plane. Additionally, measures of altered venous return were not assessed. Despite these limitations, systemic hypotension for a period that was likely sufficient to result in the observed pathological changes was not identified. It seems most likely that the factors that altered spinal cord blood flow and perfusion in the dogs were confined to the lumbar spinal cord region and did not affect systemic physiologic variables.
The dogs of the present report developed acute poliomyelomalacia that affected the lumbar portion of the spinal cord after undergoing decompressive surgery for IVDH. The distribution of lesions in the gray matter observed by MRI mirrored the distribution of the gray matter changes observed microscopically. Although the physiologic variables measured in the dogs during anesthesia were not considered sufficiently negative to have detrimentally impacted or altered spinal cord perfusion or blood flow, the microscopic findings in the gray matter of the lumbar intumescence were consistent with ischemia. Clinicians should consider spinal cord ischemia as a potential cause for a dramatic postoperative decline in neurologic function in dogs that have undergone anesthesia for MRI and decompressive surgery for IVDH.
ABBREVIATIONS
| DAP | Diastolic arterial blood pressure |
| ETco2 | End-tidal carbon dioxide concentration |
| FCEM | Fibrocartilaginous embolic myelopathy |
| IVD | Intervertebral disk |
| IVDH | Intervertebral disk herniation |
| MAP | Mean arterial blood pressure |
| SAP | Systolic arterial blood pressure |
Footnotes
Vetsulin, Intervet Inc (Merck Animal Health), Madison, NJ.
Magnitude 3150, Invivo Research Inc, Orlando, Fla.
Medrad Veris, Bayer HealthCare LLC, Whippany, NJ.
3.0T Skyra, Siemens Medical Solutions USA Inc, Malvern, Pa.
Magnivest, Bayer HealthCare LLC, Whippany, NJ.
References
1. Freer SR, Scrivani PV. Postoperative susceptibility artifact during magnetic resonance imaging of the vertebral column in two dogs and a cat. Vet Radiol Ultrasound 2008;49:30–34.
2. DeGirolami U, Zivin JA. Neuropathology of experimental spinal cord ischemia in the rabbit. J Neuropathol Exp Neurol 1982;41:129–149.
3. Yamamoto N, Takano H, Kitagawa H, et al. Changes of evoked action potentials and histology of the spinal cord, and hind limb dysfunction in spinal cord ischemia of cats. J Spinal Disord 1994;:285–295.
4. Ragle C, Baetge C, Yiannikouris S, et al. Development of equine post anaesthetic myelopathy: thirty cases (1979–2010). Equine Vet Educ 2011;23:630–635.
5. Gold JR, Divers TJ, Lahunta A. Postanesthetic poliomyelopathy in a 7-day-old calf. J Vet Intern Med 2005;19:775–778.
6. Kim SW, Kim RC, Choi BH, et al. Non-traumatic ischaemic myelopathy: a review of 25 cases. Paraplegia 1988;26:262–272.
7. Alisauskaite N, Spitzbarth I, Baumgartner W, et al. Chronic post-traumatic intramedullary lesions in dogs, a translational model. PLoS One 2017;12:e0187746.
8. De Risio L, Platt SR. Fibrocartilaginous embolic myelopathy in small animals. Vet Clin North Am Small Anim Pract 2010;40:859–869.
9. Summers BA, Cummings JF, DeLahunta A. Degenerative diseases of the central nervous system. In: Veterinary neuropathology. St Louis: Mosby, 1995;208–350.
10. Griffiths IR. Some aspects of the pathology and pathogenesis of the myelopathy caused by disc protrusions in the dog. J Neurol Neurosurg Psychiatry 1972;35:403–413.
11. Henke D, Gorgas D, Doherr MG, et al. Longitudinal extension of myelomalacia by intramedullary and subdural hemorrhage in a canine model of spinal cord injury. Spine J 2016;16:82–90.
12. Henke D, Vandevelde M, Doherr MG, et al. Correlations between severity of clinical signs and histopathological changes in 60 dogs with spinal cord injury associated with acute thoracolumbar intervertebral disc disease. Vet J 2013;198:70–75.
13. Smith PM, Jeffery ND. Histological and ultrastructural analysis of white matter damage after naturally-occurring spinal cord injury. Brain Pathol 2006;16:99–109.
14. Lewis MJ, Cohen EB, Olby NJ. Magnetic resonance imaging features of dogs with incomplete recovery after acute, severe spinal cord injury. Spinal Cord 2018;56:133–141.
15. Turnbull IM. Microvasculature of the human spinal cord. J Neurosurg 1971;35:141–147.
16. Acker G, Schneider UC, Grozdanovic Z, et al. Cervical disc herniation as a trigger for temporary cervical cord ischemia. J Spine Surg 2016;2:135–138.
17. Kato S, Kawahara N, Tomita K, et al. Effects on spinal cord blood flow and neurologic function secondary to interruption of bilateral segmental arteries which supply the artery of Adamkiewicz: an experimental study using a dog model. Spine (Phila Pa 1976) 2008;33:1533–1541.
18. de Lahunta A, Glass EN, Kent M. Lower motor neuron: spinal nerve general somatic efferent system. In: Veterinary neuroanatomy and clinical neurology. 4th ed. St Louis: Elsevier, 2015;162–196.
19. Etz CD, Luehr M, Kari FA, et al. Paraplegia after extensive thoracic and thoracoabdominal aortic aneurysm repair: does critical spinal cord ischemia occur postoperatively? J Thorac Cardiovasc Surg 2008;135:324–330.
20. Blakemore WF, Jefferies A, White RA, et al. Spinal cord malacia following general anaesthesia in the horse. Vet Rec 1984;114:569–570.
21. Gaynor JS, Dunlop CI, Wagner AE, et al. Complications and mortality associated with anesthesia in dogs and cats. J Am Anim Hosp Assoc 1999;35:13–17.
22. Iizuka T, Kamata M, Yanagawa M, et al. Incidence of intraoperative hypotension during isoflurane-fentanyl and propofolfentanyl anaesthesia in dogs. Vet J 2013;198:289–291.
23. 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.
24. Fenn J, Laber E, Williams K, et al. Associations between anesthetic variables and functional outcome in dogs with thoracolumbar intervertebral disk extrusion undergoing decompressive hemilaminectomy. J Vet Intern Med 2017;31:814–824.
25. Davis GJ, Brown DC. Prognostic indicators for time to ambulation after surgical decompression in nonambulatory dogs with acute thoracolumbar disk extrusions: 112 cases. Vet Surg 2002;31:513–518.
26. Dixon A, Fauber AE. Effect of anesthesia-associated hypotension on neurologic outcome in dogs undergoing hemilaminectomy because of acute, severe thoracolumbar intervertebral disk herniation: 56 cases (2007–2013). J Am Vet Med Assoc 2017;250:417–423.
27. Kise Y, Kuniyoshi Y, Inafuku H, et al. Directly measuring spinal cord blood flow and spinal cord perfusion pressure via the collateral network: correlations with changes in systemic blood pressure. J Thorac Cardiovasc Surg 2015;149:360–366.
28. Griffiths IR, Pitts LH, Crawford RA, et al. Spinal cord compression and blood flow. I. The effect of raised cerebrospinal fluid pressure on spinal cord blood flow. Neurology 1978;28:1145–1151.
29. Noussitou FL, Gorgas D, Rohrbach H, et al. Assessment of intramedullary spinal pressure in small breed dogs with thoracolumbar disk extrusion undergoing hemilaminectomy. Vet Surg 2015;44:944–948.
30. Langerhuus L, Miles J. Proportion recovery and times to ambulation for non-ambulatory dogs with thoracolumbar disc extrusions treated with hemilaminectomy or conservative treatment: a systematic review and meta-analysis of caseseries studies. Vet J 2017;220:7–16.
31. Dhupa S, Glickman N, Waters DJ. Reoperative neurosurgery in dogs with thoracolumbar disc disease. Vet Surg 1999;28:421–428.
32. 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.
33. Ruddle TL, Allen DA, Schertel ER, et al. Outcome and prognostic factors in non-ambulatory Hansen type I intervertebral disc extrusions: 308 cases. Vet Comp Orthop Traumatol 2006;19:29–34.
34. Blight AR. Cellular morphology of chronic spinal cord injury in the cat: analysis of myelinated axons by line-sampling. Neuroscience 1983;10:521–543.
35. Olby N, Levine J, Harris T, et al. Long-term functional outcome of dogs with severe injuries of the thoracolumbar spinal cord: 87 cases (1996–2001). J Am Vet Med Assoc 2003;222:762–769.
36. Jeffery ND, Barker AK, Hu HZ, et al. Factors associated with recovery from paraplegia in dogs with loss of pain perception in the pelvic limbs following intervertebral disk herniation. J Am Vet Med Assoc 2016;248:386–394.
