Surgical Removal of an Intramedullary Spinal Cord Foreign Body Granuloma in a Dog
A 2-year-old, spayed female, mixed-breed dog was presented for evaluation of a progressive asymmetric tetraparesis and cranial nerve deficits with a 3-week duration. Computed tomography showed a contrast-enhancing lesion along the left side of the junction of the medulla and the cervical spinal cord. An exploratory surgery determined the presence of an intramedullary lesion of the first cervical spinal cord segment. The mass was removed through a dorsal midline myelotomy. Microscopic examination identified a foreign body granuloma that contained a dense, anisotropic outer wall, supporting the conclusion that the foreign body was of plant origin. The dog recovered to a more improved ambulatory status than prior to surgery.
Case Report
An 11-kg, 2-year-old, spayed female, mixed-breed dog was referred to the Texas A&M University Veterinary Medical Teaching Hospital (TAMU-VMTH) with a 3-week history of cervical spinal hyperesthesia, lethargy, and progressive asymmetric tetraparesis that was worse on the left side. Treatment instituted before referral included one injection of dexamethasone (0.15 mg/kg body weight, intramuscularly [IM]) and flunixin meglumine (1 mg/kg body weight, IM). Initially, the clinical signs improved but then progressed to include a right-sided curvature of the neck (i.e., torticollis) and a left hemiparesis. Dexamethasone treatment was repeated with little improvement of clinical signs. The dog began to lose the ability to control urination and bowel movements, and the right ear was drooped. The dog lived indoors and outdoors in a fenced yard with one other dog.
Physical examination revealed a thin body condition. On palpation of the cranium, there was temporalis and masseter muscle atrophy on the left side. On neurological examination, the dog was alert with severe, left-sided hemiparesis and a torticollis to the right. Conscious proprioception was absent in the left thoracic and pelvic limbs. Postural reaction testing defined mild deficits in the right limbs and severe deficits in the left limbs. Cranial nerve examination defined a decreased menace response with intact pupillary light reflexes in the right eye, decreased palpebral reflex on the right side, hypalgesia on the right side of the face, and a lip and ear droop on the left side of the face. Spinal reflexes were intact. A painful response was elicited on extended and flexed manipulations of the cervical spine. Based on the neurological examination, a multifocal central nervous system disorder affecting the brain stem and cervical spinal cord cranial to the cervical intumescence was suspected. The absent menace response with intact pupillary light reflexes and depressed palpebral reflex indicated a right facial nerve deficit. Atrophy of the temporalis and masseter muscles and decreased facial sensation were suggestive of deficits localized to the motor and sensory branches of the trigeminal nerve. Results from a complete blood cell count, the serum biochemical profile, and urinalysis were within reference ranges. Computed tomographya (CT) of the brain and cranial region of the cervical spinal cord was performed under general anesthesia. Computed tomographic images consisted of 3-mm thick, contiguous transverse and dorsal sections of the brain. Images were obtained before and after intravenous (IV) contrast medium injection.b A contrast-enhancing lesion measuring 1.8 × 1 × 1 cm was present along the left side of the junction of the medulla oblongata and the cervical spinal cord [Figures 1A–1C]. The presumptive diagnosis was a neoplasm, with some consideration given to an inflammatory lesion (i.e., granulomatous meningoen-cephalomyelitis). Attempt to collect cerebrospinal fluid (CSF) from the caudal lumbar region (e.g., between the fifth and sixth lumbar vertebrae) was unsuccessful. Serological results were negative for Coccidioides immitis, Cryptococcus neoformans, Toxoplasma gondii, and Neospora caninum. Three days later, an exploratory surgery of the mass was performed in an attempt to obtain a diagnosis and allow for spinal cord decompression.
Surgical Procedure
The dog was premedicated with a combination of glycopyrrolate (0.01 mg/kg body weight, IM) and buprenorphine hydrochloride (15 μg/kg body weight, IM). Anesthesia was induced with propofol (3 mg/kg body weight, IV), followed by endotracheal intubation, and maintained with 2.0% isoflurane in oxygen.
The dog was positioned in sternal recumbency with the head in a 45° flexed position. A dorsal approach to the cranial cervical spine and suboccipital approach to the cranium were performed using previously described techniques.12 The atlanto-occipital membrane was excised to expose the foramen magnum and the dura mater. A pneumatic drill was used to enlarge the foramen magnum 1.0 cm further in diameter. The left and right halves of the neural arch of the first cervical vertebra (C1) were partially removed using a pneumatic drill and rongeurs. The rostral half of the spinous process of the second cervical vertebra (C2) was removed with bone-cutting forceps and pneumatic drill. The left half of the neural arch of C2 was subsequently removed. Epidural fat was absent from the region of the C1 and C2 spinal cord segments. The dura mater was incised over the dorsal midline of the C1 spinal cord segment and extended cranially to expose the caudal vermis of the cerebellum. Several stay sutures were preplaced in the dura mater for retraction and exposure of the spinal cord. Gentle palpation of the C1 spinal cord segment revealed a firm mass in the dorsal area on the left side. Cytopathological examination of a fine-needle aspirate of the mass was nondiagnostic. On midline myelotomy of C1, there was a gray and tan, friable mass. A plane of dissection between the mass and normal tissue was identified, and the mass was removed using gentle traction with microdissection instruments. A section of the mass was obtained for immediate intraoperative cytopathological examination. The remainder of the mass was placed in neutral buffered formalin for histopathological evaluation. Impression smears from the mass revealed a mixture of many nondegenerate neutrophils and heavily vacuolated macrophages with fewer small lymphocytes, neurons, and plasma cells. No etiological agents or neoplastic cells were seen. The cytopathological pattern was suggestive of pyogranulomatous inflammation. Most of the mass was removed, and the spinal cord and surrounding tissues were copiously lavaged with saline (0.9% NaCl) solution. Normal spinal cord tissue and focal areas of myelomalacia were visualized. The incision was closed in a routine manner.
Perioperative medication included administration of methylprednisolone sodium succinate (30 mg/kg body weight, IV) at the start of surgery and another dose (15 mg/kg body weight, IV) 2 hours later. Antibiotic treatment included cephazolin sodium (22 mg/kg body weight, IV, q 2 hours intraoperatively and then q 8 hours for 24 hours). Pain was controlled with buprenorphine (10 μg/kg body weight, IM, q 12 hours). Postoperatively there was no respiratory compromise, and the dog still had good superficial pain perception as well as some voluntary motor activity in the pelvic limbs. The thoracic limbs showed severe spastic rigidity that was exacerbated with external stimuli.
Postoperatively the dog was administered prednisone (1 mg/kg body weight, per os [PO], q 12 hours) based on the intraoperative cytopathology report of chronic suppurative inflammation with no evidence of neoplastic cells or infectious agents. Clindamycin (10 mg/kg body weight, PO, q 12 hours) was administered until serological testing for protozoal organisms was complete.
On microscopic examination of the tissue removed at surgery, there was a prominent, chronic-active inflammation characterized by focal and diffuse accumulations of macrophages, lymphocytes, plasma cells, and neutrophils. There also was focal and diffuse fibrous connective tissue formation. Identifiable central nervous system (CNS) tissue consisting of a narrow band of white matter that formed the outer margin along one border of the specimen, plus small islands of gray matter containing isolated neuronal cell bodies, confirmed that the spinal cord was involved, even though the specific extent of involvement could not be determined. Although no recognizable meningeal tissue was detected, the marked focal fibrosis in the lesion suggested meningeal involvement.
At discharge, 8 days after surgery, the dog could stand with support but was severely tetraparetic and nonambulatory. Fifty-two days after discharge, the dog returned for a recheck evaluation. The owner reported that the dog was able to stand unassisted 3 weeks after the surgery. At 5 weeks after the surgery, the dog was able to walk. Neurological examination revealed mild to moderate tetraparesis and a right-sided torticollis. The rest of the neurological examination was the same as before surgery. The brain stem and cranial cervical spinal cord were reexamined by CT. There was no evidence of a mass on pre- or postcontrast images [Figures 1D–1F]. The dog was continued on the same dose of prednisone (1 mg/kg body weight, PO, q 12 hours) for an additional 10 days before the dose was decreased to anti-inflammatory levels (0.5 mg/kg body weight, PO, q 12 hours). The glucocorticoid dose was reduced and discontinued after 4 months of therapy. At the time of manuscript submission, the dog had continued to regain motor strength and the torticollis had almost resolved.
Seven months after the initial presentation, additional sections of the mass were cut in a continued attempt to determine a definitive diagnosis. On microscopic examination at this time, there was a foreign body in the center of the tissue specimen that had an ill-defined center surrounded by a dense, anisotropic outer wall. The positive anisotropism, plus the cellular characteristics of the outer wall, supported the conclusion that the foreign body was of plant origin [Figures 2A, 2B]. No organisms were detected following examination with special stains.
Discussion
Plant materials such as grass awns have been the most widely associated foreign bodies in animals, and their migratory capacity can precipitate a variety of syndromes.3–6 The barbed nature of an awn prevents retrograde migration and causes subsequent forward migration to occur with any body movement.3 Common portals of entry are the external ear canal, interdigital web, conjunctiva, third eyelid, and the upper respiratory and gastrointestinal tracts.347–9 Geographic regional differences influence the specific causative agent.10
Common neurological diseases in dogs and cats associated with foreign body migration include encephalitis1112 and diskospondylitis6 or vertebral osteomyelitis1314 involving the cranial lumbar area (i.e., second to fourth lumbar vertebral region). This occurs subsequent to migration of the awn through the lung parenchyma, mediastinum and crura of the diaphragm, or gastrointestinal tract to the paravertebral structures.3131516 Foreign body-induced diskospondylitis in a dog has also been reported to be caused by a porcupine quill migration.17 An epidural injection has been associated with an epidural abscess and diskospondylitis in a dog.18
Various organisms have been isolated from dogs and cats following foreign body migration. Streptococcus spp. was most frequently isolated, followed less frequently by Staphylococcus aureus, Pasteurella multocida, Actinomyces spp., and Nocardia spp.3 In contrast, other reports indicate that Actinomyces spp. is a common pathogen found in association with migration of the grass awns.89 Unfortunately, bacterial and fungal cultures were not performed in the present case. Even though no organisms were detected microscopically in the tissue examined, it was still suspected that organisms such as bacteria could have accompanied the traumatic migration of the foreign body into the affected spinal cord.
Although foreign body migration to the cervical vertebral spine of dogs has been reported as being unusual,1617 migration to involve the spinal cord parenchyma as described in the present case has not been previously reported. The exact mechanism by which the migration to the cord occurred in this case could not be determined, but some possibilities were considered. The plant material could have entered C1 spinal cord segment at the large space between the occiput and atlas either dorsally or more likely ventrally, subsequent to perforation through the pharyngeal soft tissues.1116
In humans, foreign body migration in the CNS is considered rare, but when it occurs there are serious consequences. Granulomatous reactions to foreign bodies have been reported as sequelae following use of suture material,1920 hemostatic materials,21 cotton pads,22–25 cadaveric26 or synthetic dural grafts,27 and shunt catheters.2829 Spontaneous migration of metallic foreign bodies such as bullets and hemostatic clips has also been described.30 Circulation of CSF has been proposed to facilitate migration of hemostatic clips out of the ventricles into the subarachnoid space, followed by gravity-induced migration to the lumbar region where lesions occurred.30
Intramedullary spinal cord abscess is rare in humans. It usually occurs in the setting of systemic infection with multiple septic foci elsewhere.3132 Among other routes, abscesses may arise from a remote site by hematogenous spread, contiguity (i.e., trauma, dermal sinus, stab wound), or lymphatics.31–34 Bacteria commonly associated with intramedullary spinal cord abscesses are Staphylococcus spp., Streptococcus spp., Gram-negative, and Listeria spp.3334 In some reports, most cultures yielded no growth.3234 Prognostic features of spinal cord abscess that correlated with good outcome were an age <25 years and duration of severe motor loss <3 days.32 Management of an intramedullary spinal cord abscess is directed at prompt decompressive laminectomy, myelotomy, and surgical drainage, along with appropriate antimicrobial therapy.31–34 Corticosteroid therapy is recommended for the treatment of vasogenic edema and any mass effect. Because corticosteroid therapy may decrease the penetration of antibiotics into CNS tissue, it should be discontinued when the edema and mass effect improve or resolve.35 Immunosuppressive corticosteroid treatment in this dog could have been detrimental, had the foreign body not been walled-off by the granuloma and completely surgically excised.
Conclusion
Surgical exploration in this case was associated with a positive outcome. Removal of the granuloma alleviated the mass effect and established the diagnosis. Although the surgical procedure caused a transient worsening of neurological signs, the dog eventually recovered to a more improved ambulatory status than prior to surgery.
GE Sytec 400; General Electric, Milwaukee, WI
Iothalamate sodium; Conray 400 Mallinckrodt Inc., St. Louis, MO
Acknowledgment
The authors gratefully acknowledge Dr. Robert Dunstan for his assistance with the photomicrographs.
Citation: Journal of the American Animal Hospital Association 39, 3; 10.5326/0390306
Citation: Journal of the American Animal Hospital Association 39, 3; 10.5326/0390306
Postcontrast transverse computed tomographic images of the caudal brain stem and cranial cervical spinal cord of a 2-year-old, mixed-breed dog with evidence of multifocal neurological disease before (1A, 1B, 1C) and after (1D, 1E, 1F) mass removal. Note the contrast enhancement of the lesion along the left side (arrow) of the junction of the medulla oblongata with the cervical spinal cord before mass removal and its absence at the 2-month recheck.
Histopathology of the mass identified in the dog from Figure 1. (2A) Spinal cord, first cervical level. A prominent, elongated foreign body with an ill-defined center and recognizable outer wall is surrounded by a prominent inflammation characterized by accumulation of macrophages, lymphocytes, plasma cells, neutrophils, and fibrous connective tissue formation (Hematoxylin and eosin stain, 10×; bar=0.100 mm). (2B) Higher magnification of Figure 2A. The upper outer wall of the foreign body (extending through the center of the illustration) consists of very large cells with prominent cell walls compatible with plant tissue. The less well-defined central part of the structure (lower part of illustration) contains scattered neutrophils and a few plasma cells. Inflammation as described in Figure 2A is present in the upper part of the illustration (Hematoxylin and eosin stain, 40×; bar=0.025 mm).
Contributor Notes
