Introduction
A 3-year-old 31.1-kg castrated male mixed-breed dog was presented to the University of Illinois Neurology and Neurosurgery Service for evaluation because of a 1- to 2-week history of progressive paraparesis, knuckling of the hind feet, and difficulty posturing to urinate or defecate. When posturing to defecate, the dog would often fall to the side and stop defecating. The referring veterinarian advised the owner to restrict the dog’s activity and prescribed gabapentin (13 mg/kg, PO, q 12 h), carprofen (2.4 mg/kg, PO, q 12 h), and methocarbamol (16 mg/kg, PO, q 12 h). The dog’s signs of pain minimally improved with the medications, and despite limited activity, the owner reported that the dog’s mobility worsened and its appetite decreased. The clinical signs had not been observed prior to this episode, and there was no other pertinent medical history.
On physical examination, the dog was bright, alert, and responsive, with vital signs (heart rate, respiratory rate, and rectal temperature) within the respective reference intervals. The dog tensed on abdominal palpation, and its urinary bladder was large and firm. Manual expression of the bladder was difficult, and the dog was hyperreactive to touch near the tail base and anus. Results of a rectal examination were unremarkable. On neurologic examination, the dog was mentally appropriate with a kyphotic posture, had difficulty using its pelvic limbs when rising from a slippery surface, and was paraparetic but weakly ambulatory with a short-strided gait for the pelvic limbs. The cranial nerve responses were appropriate; postural reactions of the thoracic limbs were intact, but paw placing was delayed for the pelvic limbs. Withdrawal reflexes were normal for all limbs, and the cutaneous trunci and perineal reflexes were intact. The patellar reflex of the right pelvic limb was mildly decreased, and muscle tone in the pelvic limbs was increased bilaterally. Hyperesthesia was evident on paraspinal palpation of the lumbar region. Most clinical signs were consistent with a lesion in the T3-L3 spinal cord segments; however, the decreased patellar reflex potentially suggested involvement of more caudal spinal cord segments. Results of a CBC and serum biochemical analysis were within the respective reference intervals. The patient was hospitalized; a continuous rate infusion of hydromorphone hydrochloride (0.01 mg/kg/h) and gabapentin (9.6 mg/kg, PO, q 8 h) treatment was administered.
In descending order of likelihood, the initial differential diagnoses included intervertebral disc disease, neoplasia, immune-mediated inflammatory disease, and infectious disease (eg, neosporosis, toxoplasmosis, diskospondylitis or empyema, or fungal infection). The dog was anesthetized, and MRI of the lumbar vertebral column was performed with a 3T scanner (Skyra; Siemens Medical Solutions). The MRI protocol included sagittal half-Fourier single-shot turbo spin-echo, T2-weighted (T2W) and T1-weighted (T1W) pre- and postcontrast, dorsal T2W-STIR, T1W pre- and postcontrast fat-saturated, and transverse T2W, T2*-weighted (T2*W), T2W-FLAIR, and T1W pre- and postcontrast sequences. The postcontrast images were acquired after IV administration of dimeglumine gadopentetate (0.2 mL/kg).
A well-circumscribed, centrally located oval intramedullary mass (approx 5 × 1.2 × 0.8 cm) occupied almost the entire diameter of the spinal cord from the cranial aspect of L4 to the middle of L5, with complete attenuation of CSF flow at the dorsal aspect of the spinal cord. The mass was hyperintense on T2W, STIR, and FLAIR images and isointense on T1W images; contrast enhancement was strong and homogenous. Well-circumscribed regions of hypointensity (ie, a cap sign) were observed in contact with the cranial and caudal margins of the mass on T2W images (Figures 1 and 2). These regions demonstrated susceptibility artifact on T2*W images and were mildly hypointense on T1W images, suggesting chronic hemorrhage. Moderate patchy lesions that were hyperintense on T2W, FLAIR, and STIR images and non–contrast enhancing were present within the spinal cord cranial and caudal to the mass, suggestive of vasogenic edema. Given the lesion’s characteristics, the primary differential diagnosis was an ependymoma. Other differential diagnoses included other glial cell neoplasia (eg, astrocytoma), atypical nephroblastoma, lymphoma, histiocytic sarcoma, and granuloma.
Representative sagittal T2-weighted (T2W; A) and T1-weighted (T1W) postcontrast (B) images and transverse T2W (C), T1W precontrast (D), and postcontrast (E) images obtained by MRI of the lumbar vertebral column of a 3-year-old mixed-breed dog that was evaluated because of paraparesis, knuckling of the hind feet, and difficulty posturing to urinate or defecate. A well-circumscribed oval intramedullary mass is present from the cranial aspect of L4 to the middle of L5. The mass occupies almost the entire diameter of the spinal cord, completely attenuates CSF flow at the dorsal aspect of the spinal cord, and is markedly hyperintense on the T2W images and isointense on the precontrast T1W image with subsequent strong, homogeneous contrast enhancement. Moderate patchy lesions within the spinal cord cranial and caudal to the mass (arrows) are hyperintense and non–contrast enhancing on the T2W and T1W sagittal images, respectively, suggesting vasogenic edema (A and B). Hypointense regions in contact with the cranial and caudal aspects of the mass are evident on the T2W sagittal image (A).
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Representative sagittal T2-weighted (T2W; A) and T1-weighted (T1W) postcontrast (B) images and transverse T2W (C), T1W precontrast (D), and postcontrast (E) images obtained by MRI of the lumbar vertebral column of a 3-year-old mixed-breed dog that was evaluated because of paraparesis, knuckling of the hind feet, and difficulty posturing to urinate or defecate. A well-circumscribed oval intramedullary mass is present from the cranial aspect of L4 to the middle of L5. The mass occupies almost the entire diameter of the spinal cord, completely attenuates CSF flow at the dorsal aspect of the spinal cord, and is markedly hyperintense on the T2W images and isointense on the precontrast T1W image with subsequent strong, homogeneous contrast enhancement. Moderate patchy lesions within the spinal cord cranial and caudal to the mass (arrows) are hyperintense and non–contrast enhancing on the T2W and T1W sagittal images, respectively, suggesting vasogenic edema (A and B). Hypointense regions in contact with the cranial and caudal aspects of the mass are evident on the T2W sagittal image (A).
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Representative sagittal T2-weighted (T2W; A) and T1-weighted (T1W) postcontrast (B) images and transverse T2W (C), T1W precontrast (D), and postcontrast (E) images obtained by MRI of the lumbar vertebral column of a 3-year-old mixed-breed dog that was evaluated because of paraparesis, knuckling of the hind feet, and difficulty posturing to urinate or defecate. A well-circumscribed oval intramedullary mass is present from the cranial aspect of L4 to the middle of L5. The mass occupies almost the entire diameter of the spinal cord, completely attenuates CSF flow at the dorsal aspect of the spinal cord, and is markedly hyperintense on the T2W images and isointense on the precontrast T1W image with subsequent strong, homogeneous contrast enhancement. Moderate patchy lesions within the spinal cord cranial and caudal to the mass (arrows) are hyperintense and non–contrast enhancing on the T2W and T1W sagittal images, respectively, suggesting vasogenic edema (A and B). Hypointense regions in contact with the cranial and caudal aspects of the mass are evident on the T2W sagittal image (A).
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Transverse MRI images obtained cranial (A and B) and caudal (C and D) to the mass in Figure 1. A and C—T2W images. B and D—T2*-weighted images. There are well-circumscribed regions of hypointensity in the T2W images and signal dropout on T2*-weighted images (arrowheads), indicating hemorrhage.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Transverse MRI images obtained cranial (A and B) and caudal (C and D) to the mass in Figure 1. A and C—T2W images. B and D—T2*-weighted images. There are well-circumscribed regions of hypointensity in the T2W images and signal dropout on T2*-weighted images (arrowheads), indicating hemorrhage.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Transverse MRI images obtained cranial (A and B) and caudal (C and D) to the mass in Figure 1. A and C—T2W images. B and D—T2*-weighted images. There are well-circumscribed regions of hypointensity in the T2W images and signal dropout on T2*-weighted images (arrowheads), indicating hemorrhage.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
On the basis of MRI findings, CSF was collected from the lumbar cistern for analysis. The nucleated cell count was 81 cells/mm3 (reference interval,3 < 5 cells/mm), RBC count was 523 cells/mm3, and total protein concentration was 536.1 mg/dL (reference interval, < 45 mg/dL). The nucleated cells included large mononuclear cells (56%), small mononuclear cells (39%), and nondegenerate neutrophils (5%), which was consistent with mononuclear pleocytosis. Given the location of the mass at L4-L5 and the concern for neoplasia, fine-needle aspiration of the mass was performed in the same manner as collection of CSF from the lumbar cistern during the same anesthetic episode; aspirates were submitted for cytologic evaluation.
On cytologic examination, the samples were highly cellular and consisted of many variably cohesive clusters of round to cuboidal cells with rare columnar epithelial cells (Figure 3). These cells had round or rarely indented nuclei with stippled to lacy chromatin, occasional small nucleoli, and a small to moderate amount of eosinophilic, finely granular cytoplasm. Anisocytosis and anisokaryosis were minimal to mild, and acinar-like formations of cells were occasionally noted. Intercellular junctions between these cells were often indistinct, resulting in a neuroendocrine-like appearance. Low numbers of vascular structures were also identified in close association with large clusters of these cells. The sample was interpreted as consistent with primary CNS neoplasia. The cytologic features were most compatible with an ependymoma, although astrocytoma or nephroblastoma were also considered. Meningioma could not be completely excluded on the basis of cytologic features but was considered unlikely, given the intramedullary location of the mass. Combining the cytologic findings with the MRI results, the ependymoma was considered the most likely diagnosis. The dog recovered uneventfully from anesthesia and the previous treatments were continued.
Representative photomicrographs of a fine-needle aspirate sample obtained from the mass in Figure 1. Large clusters of round to cuboidal cells are present in close association with a vascular structure (asterisk; A) and in acinar-like formations (daggers; B). Inset—Enlargement of cells in panel B showing cellular detail. Wright-Giemsa stain; bar = 10 µm.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Representative photomicrographs of a fine-needle aspirate sample obtained from the mass in Figure 1. Large clusters of round to cuboidal cells are present in close association with a vascular structure (asterisk; A) and in acinar-like formations (daggers; B). Inset—Enlargement of cells in panel B showing cellular detail. Wright-Giemsa stain; bar = 10 µm.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Representative photomicrographs of a fine-needle aspirate sample obtained from the mass in Figure 1. Large clusters of round to cuboidal cells are present in close association with a vascular structure (asterisk; A) and in acinar-like formations (daggers; B). Inset—Enlargement of cells in panel B showing cellular detail. Wright-Giemsa stain; bar = 10 µm.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Representative photomicrographs of a fine-needle aspirate sample obtained from the mass in Figure 1. Large clusters of round to cuboidal cells are present in close association with a vascular structure (asterisk; A) and in acinar-like formations (daggers; B). Inset—Enlargement of cells in panel B showing cellular detail. Wright-Giemsa stain; bar = 10 µm.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Representative photomicrographs of a fine-needle aspirate sample obtained from the mass in Figure 1. Large clusters of round to cuboidal cells are present in close association with a vascular structure (asterisk; A) and in acinar-like formations (daggers; B). Inset—Enlargement of cells in panel B showing cellular detail. Wright-Giemsa stain; bar = 10 µm.
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Owing to the location and size of the mass, complete surgical removal was unlikely to be successful; therefore, clinical recurrence would have remained a concern. Radiation treatment was discussed with the client, but euthanasia was elected because of the poor prognosis and signs of severe pain. The patient was euthanized within a few hours after the neurodiagnostic workup was performed, and a postmortem examination of the spinal cord was performed. A distinct mass was not appreciated on gross examination, but the areas around the L4-L5 spinal cord segments were slightly bulging, malacic, and mildly reddened with hemorrhage. The L4-L5 and adjacent lumbar spinal cord regions were fixed in neutral-buffered 10% formalin solution, and multiple cross sections were submitted for histologic evaluation. A densely cellular, nonencapsulated, partially circumscribed neoplasm was found to expand and efface the gray matter and central canal, extend to the white matter, and compress the adjacent neuropil. The neoplasm was composed of round to polygonal cells arranged in solidly cellular areas and rosette-like structures around a central vascular space (ie, pseudorosettes) on a fine fibrovascular stroma. The neoplastic cells had indistinct cell borders, scant to moderate amounts of eosinophilic fibrillar cytoplasm, and round to oval nuclei with finely stippled to dense chromatin and single nucleoli (Figure 4). Mild anisocytosis and anisokaryosis were present as well as 6 mitotic figures/10 hpf (400×). The neoplasm extended to but did not infiltrate the meninges of several sections and encompassed up to 80% of the neuropil in some sections, with blending of neoplastic cells into the adjacent parenchyma. There was equivocal staining of phosphotungstic acid-hematoxylin suggesting the presence of basal bodies. There was no reactivity on immunohistochemical staining for glial fibrillary acidic protein (GFAP). Although true rosettes were not present, these histologic findings were more suggestive of an ependymoma than an astrocytoma. Hemorrhage was noted grossly and was confirmed on MRI by histologic examination of samples obtained cranial and caudal to the mass. Perilesional edema was not confirmed, but areas of malacia were present surrounding the cranial and caudal aspects of the mass.
Photomicrographs of the spinal cord showing the neoplasm in Figure 1 at the level of the L4-L5 spinal cord segments. A and B—A partially circumscribed, highly cellular, expansile neoplasm is present, expanding and effacing the gray matter and central canal and extending to the white matter with a blending of neoplastic cells into the adjacent parenchyma, thereby making distinction of the margins difficult (arrows). H&E stain; bar = 500 µm. C and D—The mass is composed of round to polygonal cells with basally located nuclei arranged in solidly cellular areas and perivascular pseudorosettes (asterisks). The nuclei are round to oval with finely stippled to dense chromatin and single nucleoli. There are frequent cell-free pale eosinophilic zones throughout the neoplasm (plus sign). H&E stain; bar = 500 µm (A and B), 50 µm (C), and 20 µm (D).
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Photomicrographs of the spinal cord showing the neoplasm in Figure 1 at the level of the L4-L5 spinal cord segments. A and B—A partially circumscribed, highly cellular, expansile neoplasm is present, expanding and effacing the gray matter and central canal and extending to the white matter with a blending of neoplastic cells into the adjacent parenchyma, thereby making distinction of the margins difficult (arrows). H&E stain; bar = 500 µm. C and D—The mass is composed of round to polygonal cells with basally located nuclei arranged in solidly cellular areas and perivascular pseudorosettes (asterisks). The nuclei are round to oval with finely stippled to dense chromatin and single nucleoli. There are frequent cell-free pale eosinophilic zones throughout the neoplasm (plus sign). H&E stain; bar = 500 µm (A and B), 50 µm (C), and 20 µm (D).
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Photomicrographs of the spinal cord showing the neoplasm in Figure 1 at the level of the L4-L5 spinal cord segments. A and B—A partially circumscribed, highly cellular, expansile neoplasm is present, expanding and effacing the gray matter and central canal and extending to the white matter with a blending of neoplastic cells into the adjacent parenchyma, thereby making distinction of the margins difficult (arrows). H&E stain; bar = 500 µm. C and D—The mass is composed of round to polygonal cells with basally located nuclei arranged in solidly cellular areas and perivascular pseudorosettes (asterisks). The nuclei are round to oval with finely stippled to dense chromatin and single nucleoli. There are frequent cell-free pale eosinophilic zones throughout the neoplasm (plus sign). H&E stain; bar = 500 µm (A and B), 50 µm (C), and 20 µm (D).
Citation: Journal of the American Veterinary Medical Association 259, 12; 10.2460/javma.20.09.0541
Discussion
An ependymoma is a tumor of ependymal cells, which line the ventricular system within the CNS and regulate the permeability of CSF components.1 Very few cases have been reported in dogs, and most ependymomas are found in the brain or as intramedullary masses in the thoracolumbar region of the spinal cord.2 Primary intramedullary spinal cord tumors are extremely rare in people, comprising 20% to 30% of primary spinal cord tumors, of which ependymomas constitute only 30% to 40%.3 In people, ependymomas are more commonly located in the cervicothoracic spine of adults, and there is no sex predilection for development of the tumors.3 In dogs, there is no apparent breed or sex predilection, and although the number of reported cases is small, young to middle-aged dogs appear to be affected.2,4 Some human ependymomas have been associated with an autosomal dominant mutation of the neurofibromin tumor suppressor gene on chromosome 22, leading to neurofibromatosis type 2.3 The authors are unaware of any studies investigating whether genetic factors also play a role in canine ependymomas.
Ependymomas efface the surrounding tissue and may invade locally but do not typically metastasize in dogs.5 Drop metastasis has been reported in human patients with ependymomas,6 but to the authors’ knowledge, it has not been described in canine patients. Compression and displacement of surrounding nervous tissue can lead to clinical signs that vary according to tumor location, including altered mentation, cranial nerve abnormalities, tetra- or paraparesis, signs of pain, and proprioceptive deficits.7 Affected people often experience back or neck pain due to the distension of the dura and sensory or motor nerve dysfunction, depending on the location of the mass.3 Dogs have been reported to show various neurologic signs such as ataxia, paresis, and increased patellar reflexes, but have consistently had vital signs and hematologic values considered normal.7
To the authors’ knowledge, 2 publications4,7 in the veterinary literature have previously described the MRI appearance of spinal cord ependymomas in dogs. In both reports,4,7 the masses were centrally located, hyperintense on T2W images, and hypointense on T1W images. In a clinical report7 of 1 dog, the mass was not enhanced by contrast medium administration, whereas masses in 2 dogs had strong, homogeneous contrast enhancement in another investigation.4 The presence of perilesional edema or hemorrhage was not mentioned in either report. In the dog of the present report, findings on MRI indicated the mass was similar in appearance and had strong, homogeneous contrast enhancement. These same MRI features have been reported for spinal cord ependymomas in people3; furthermore, the presence of a cap sign at the upper or lower margins of the tumor (or both) on T2W images owing to hemosiderin deposition from chronic lesion hemorrhage has been described for these patients.8,9 The finding of a cap sign is more often associated with spinal cord ependymoma in people, but it can also be seen with other tumors.9 Given the appearance on the T2W, T1W, and T2*W images, the cap sign in the dog of the present report indicated chronic hemorrhage, which was confirmed on histologic examination. In people, spinal cord ependymomas commonly have perilesional edema characterized on MRI as non–contrast-enhancing areas that are hyperintense on T2W images.9 This was suspected on the basis of MRI findings for the dog of the present report but was not confirmed on histologic evaluation. Other potential causes for these patchy lesions on MRI included gliosis or myelomalacia, and the latter was confirmed in this case. Taken together, these findings suggest that ependymoma should be considered as a differential diagnosis in young or middle-aged dogs with evidence of an intramedullary mass with T2W hyperintensity, strong contrast enhancement, and a cap sign on MRI examination.
In most reported cases of ependymoma in dogs, the definitive diagnosis has been determined through gross and histologic evaluation.1,2,7,10 Gross descriptions of the masses vary and have included soft and grayish white,2 lobulated with a reddish coloration due to hemorrhage,7 and solid with a homogeneous texture.10 The histologic characteristics are somewhat inconsistent, but the most commonly described feature is the presence of rosette or pseudorosette structures.1 Human ependymomas also typically have rosettes or perivascular pseudorosettes, but a less common subtype of the tumors may not have these features.3 Multiple reports1,5 of this tumor in dogs describe cells with eosinophilic cytoplasm, dense chromatin, and staining for GFAP and vimentin.1,10 However, ependymomas in dogs can have equivocal staining for GFAP, and the lack of a positive result does not rule out the presence of this tumor type.11,12 For example, in studies by Ueno et al7 and Michimae et al,13 the ependymomas of dogs had negative results for GFAP staining, similar to the dog of the present report. Because other CNS tumors such as astrocytomas can have similar findings, it can be difficult to make a definitive diagnosis.1,14
Cytologic features of canine ependymoma have been sparsely documented in the veterinary literature, likely owing to difficulty in safely obtaining samples from the CNS. In previous reports,15,16 findings from fine-needle aspirate or smear preparations indicate highly cellular samples composed of loosely cohesive cellular aggregates sometimes surrounding blood vessels. These features were identified in the present case, as were occasional acinar-like formations of cells, which likely represented the pseudorosette formations that are consistent, but not pathognomonic, features of ependymomas.17 These findings may support a tentative diagnosis of ependymoma; however, cytologic features of choroid plexus tumors or metastatic lesions from well-differentiated adenocarcinomas may have a similar appearance,18,19 and consideration of the cytologic findings, patient signalment, and lesion localization is critical to formulating accurate differential diagnoses.
Treatment options for veterinary patients with ependymomas are limited, and the prognosis for these patients is often poor to grave. The gold-standard treatment for intramedullary spinal cord tumors in human patients is surgical removal combined with radiation therapy.7,20 Although challenging, early tumor diagnosis and surgical intervention are critical for relief of symptoms and quality of life in people.3 A positive prognostic indicator for human patients with ependymomas is the presence of a syrinx, and a negative prognostic indicator is the presence of a more malignant anaplastic ependymoma.3 Ueno et al7 described an intramedullary ependymoma located at the level of L3 in a dog, for which a dorsal laminectomy and durotomy were performed directly over the mass. Approximately 60% of the mass was removed, and the dog underwent radiation treatment at the surgical site with low-dose carboplatin administration to increase radiosensitivity. The dog returned to essentially normal function with no ataxia, signs of pain, or upper motor neuron signs and was alive 16 months after surgery, with no evidence of tumor recurrence.7 In a dog that had an intracranial clear cell ependymoma, radiation treatment over a 4-month period reduced the size of the mass but did not improve the clinical signs, which included seizures.21 Another dog underwent dorsal laminectomy, durotomy, and myelotomy for treatment of an ependymoma at the level of T13 that had caused progressive knuckling of the pelvic limbs (worse on the right side).22 The mass was separated from the spinal cord tissue by means of blunt dissection; the tumor appeared to be excised completely, and the patient regained partial function of its the pelvic limbs, with occasional dragging of the right pelvic limb, but developed fecal incontinence in the following months and was euthanized.22 A clinical report23 of multiple animals published in 2016 included a dog with an ependymoma and concurrent diastematomyelia that were treated surgically, but that patient’s neurologic status deteriorated rapidly and it was euthanized.
Documentation of imaging findings and clinical outcomes for dogs with ependymomas is important, as the number of cases in the literature is small. For the dog of the present report, a presumptive diagnosis of ependymoma was made on the basis of the patient’s age, interpretation of the MRI images, and cytologic examination results, with supporting findings on histologic evaluation. This case illustrated the potential for successful fine-needle aspiration and cytologic evaluation of a spinal cord tumor. This technique should be considered in dogs when the mass is in an accessible location as a means to acquire further diagnostic information and allow for consideration of possible treatment options.
Acknowledgments
The authors declare that there were no conflicts of interest with respect to the research, authorship, or publication of this article.
References
- 1. ↑
Miller AD, Koehler JW, Donovan TA, et al. Canine ependymoma: diagnostic criteria and common pitfalls. Vet Pathol. 2019;56(6):860–867.
- 2. ↑
de Vries-Chalmers Hoynk van Papendrecht HR, Vos JH, van Nes JJ. Spinal cord ependymoma in two young dogs. Vet Q. 1988;10(3):205–210.
- 3. ↑
Samartzis D, Gillis CC, Shih P, O'Toole JE, Fessler RG. Intramedullary spinal cord tumors: part I—epidemiology, pathophysiology, and diagnosis. Global Spine J. 2015;5(5):425–435.
- 4. ↑
Pancotto TE, Rossmeisl JH, Zimmerman K, Robertson JL, Werre SR. Intramedullary spinal cord neoplasia in 53 dogs (1990–2010): distribution, clinicopathologic characteristics, and clinical behavior. J Vet Intern Med. 2013;27(6):1500–1508.
- 5. ↑
Cordy D. Tumors of the nervous system and eye. In: Moulton J, ed. Tumours in Domestic Animals. University of California Press; 1990:645.
- 6. ↑
Rege SV, Narayan S, Patil H, Songara A. Spinal myxopapillary ependymoma with interval drop metastasis presenting as cauda equina syndrome: case report and review of literature. J Spine Surg. 2016;2(3):216–221.
- 7. ↑
Ueno H, Morimoto M, Kobayashi Y, Hizume T, Murayama N, Uzuka Y. Surgical and radiotherapy treatment of a spinal cord ependymoma in a dog. Aust Vet J. 2006;84(1–2):36–39.
- 8. ↑
Hanna MH, Bansal A, Belani P. A review of radiographic imaging findings of ependymal tumors. Neurosurg Cases Rev. 2019;2(2):028. doi: 10.23937/2643-4474/1710028
- 9. ↑
Fine MJ, Kricheff II, Freed D, Epstein FJ. Spinal cord ependymomas: MR imaging features. Radiology. 1995;197(3):655–658.
- 10. ↑
Vural SA, Besalti O, Ilhan F, Ozak A, Haligur M. Ventricular ependymoma in a German Shepherd Dog. Vet J. 2006;172(1):185–187.
- 11. ↑
Cantile C, Youssef S. Nervous system. In: Maxie M, ed. Jubb, Kennedy & Palmer's Pathology of Domestic Animals. Vol 1. Elsevier; 2015:250–406.
- 12. ↑
Higgins RJ, Bollen AW, Dickinson PJ, Sisó-Llonch S. Tumors of the nervous system. In: Meuten D, ed. Tumors in Domestic Animals. John Wiley & Sons Ltd; 2016:834–891.
- 13. ↑
Michimae Y, Morita T, Sunagawa Y, Sawada M, Okamoto Y, Shimada A. Anaplastic ependymoma in the cervical spinal cord of a Maltese dog. J Vet Med Sci. 2004;66(9):1155–1158.
- 14. ↑
Rissi DR, Barber R, Burnum A, Miller AD. Canine spinal cord glioma. J Vet Diagn Invest. 2017;29(1):126–132.
- 15. ↑
Sfacteria A, Macrì F, Perillo L, Rapisarda G, Lanteri G, Mazzullo G. Cytologic and histologic features of spinal cord ependymoma in a young dog: a case report. Vet Med (Praha). 2010;55(1):35–38.
- 16. ↑
Fernandez FR, Grindem CB, Brown TT, Sharp NJH, Saulnier M. Cytologic and histologic features of a poorly differentiated glioma in a dog. Vet Clin Pathol. 1997;26(4):182–186.
- 17. ↑
Moore A, Barger A. Central nervous system cytology. In: Barger A, MacNeil A, eds. Small Animal Cytologic Diagnosis. CRC Press; 2016:134–136.
- 18. ↑
De Lorenzi D, Mandara M. The central nervous system. In: Raskin R, Meyer D, eds. Canine and Feline Cytology: a Color Atlas and Interpretation Guide. Saunders Elsevier; 2010:358–361.
- 19. ↑
Schwartz D, Vernau W, Vernau KM. Central nervous system neoplasia in the dog and cat. In: Sharkey L, Radin M, Seelig D, eds. Veterinary Cytology. John Wiley & Sons Ltd; 2020:617–637.
- 20. ↑
Michalski J, Garcia D. Spinal canal. In: Perez C, Brady L, eds. Principles and Practice of Radiation Oncology. 3rd ed. JB Lippincott & Co; 1998:849–866.
- 21. ↑
Traslavina RP, Kent MS, Mohr FC, et al. Clear cell ependymoma in a dog. J Comp Pathol. 2013;149(1):53–56.
- 22. ↑
Jeffery ND, Phillips SM. Surgical treatment of intramedullary spinal cord neoplasia in two dogs. J Small Anim Pract. 1995;36(12):553–557.
- 23. ↑
Besalti O, Caliskan M, Can P, Vural SA, Algin O, Ahlat O. Imaging and surgical outcomes of spinal tumors in 18 dogs and one cat. J Vet Sci. 2016;17(2):225–234.
