Spinal arachnoid cysts (SACs) have been reported previously

Transcription

1 J Vet Intern Med 2002;16: Retrospective Analysis of Spinal Arachnoid Cysts in 14 Dogs Helena Rylander, David Lipsitz, Wayne L. Berry, Beverly K. Sturges, Karen M. Vernau, Peter J. Dickinson, Sonia A. Añor, Robert J. Higgins, and Richard A. LeCouteur Spinal cord dysfunction secondary to spinal arachnoid cysts (SACs) has been reported previously in dogs. This retrospective study reviews the clinical signs, radiographic findings, and outcome after surgical resection of SACs in 14 dogs. Plain vertebral column radiographs and myelography were done in all dogs. Computed tomography (CT) was done in 7 dogs and magnetic resonance (MR) imaging in 3 dogs. Affected dogs were between 1 and 12 years of age, and 8 of 14 were Rottweilers. Abnormalities detected on neurological examination depended on the location of the SAC. Five dogs had bilobed or multiple SACs. SACs were located in the cervical vertebral column in 11 dogs and in the thoracic vertebral column in 4 dogs. All dogs had dorsally or dorsolaterally located SACs. Two dogs also had additional ventrally located SACs. Spinal cord compression secondary to intervertebral disc extrusion or protrusion was demonstrated at the site of the SACs in 2 dogs. Surgical resection of the SACs was completed in all dogs. Eleven dogs were available for follow-up. Five weeks postoperatively, 7 dogs improved in neurological function, with some residual ataxia and paresis in 6 of these dogs. Neurological function had deteriorated in 4 dogs. It was concluded from this study that Rottweilers have a higher incidence of SACs than other breeds of dog. Furthermore, bilobed or multiple SACs can occur commonly, and myelography effectively localized SACs in dogs. Surgical resection of SACs resulted in improvement in neurological function in the majority of treated dogs. Key words: Canine; Myelogram; Neuroimaging; Rottweiler; Subarachnoid cyst. Spinal arachnoid cysts (SACs) have been reported previously as a cause of neurological dysfunction in dogs and cats Although most cysts are thought to be congenital, SACs have been reported in association with spinal cord trauma and intervertebral disc disease Thirty-nine cases of SACs have been reported previously In the majority of these reports, there was no age, breed, or gender predilection for cyst occurrence. In 1 report, 6 of 11 dogs were Rottweilers. 16 In most dogs, SACs were located dorsal or dorsolateral to the spinal cord, and in 4 dogs, multiple cysts were reported. In 3 dogs, multiple cysts were present at the same location and in another dog the cysts were located at 2 different anatomic sites. 9,16 Two of the dogs with multiple SACs at the same location also had a ventrally located SAC. 16 Of the 39 dogs previously reported, SACs were located only in the cervical vertebral column in 23 dogs, only in the thoracic vertebral column in 15 dogs, and in both the cervical and thoracic vertebral column in 1 dog. Myelography was used to diagnose SACs in all reports. In 9 dogs, computed tomography (CT) was done, and in 4 dogs, magnetic resonance (MR) imaging was done in addition to myelography to further evaluate the location of SACs. Surgical removal of the SAC was performed in 30 dogs. Eight dogs were euthanized after the diagnosis had From the Veterinary Medical Teaching Hospital (Rylander, Berry, Añor), the Department of Surgical and Radiological Sciences (Lipsitz, Sturges, Vernau, Dickinson, LeCouteur), the Department of Pathology, Microbiology, and Immunology (Higgins), School of Veterinary Medicine, University of California-Davis, Davis, CA. Dr Berry is presently affiliated with the Southern California Surgical Group, Irvine, CA. Dr Añor is presently affiliated with the School of Veterinary Medicine, Autonomous University of Barcelona, Barcelona, Spain. Reprint request: Helena Rylander, DVM, Veterinary Medical Teaching Hospital, University of California-Davis, One Shields Avenue, Davis, CA ; hrylander@vmth.ucdavis.edu. Submitted March 11, 2002; Revised April 30, 2002; Accepted May 20, Copyright 2002 by the American College of Veterinary Internal Medicine /02/ /$3.00/0 been made, and 1 dog was treated with an anti-inflammatory dosage of prednisone. The purpose of this retrospective study is to report the signalment, history, neurological signs, results of cerebrospinal fluid (CSF) analysis, neuroradiological and advanced imaging findings, histopathology, and outcome after surgery in 14 dogs diagnosed with SACs and to compare these findings with results of previous studies. Materials and Methods The medical records of 14 dogs with 1 or more SACs admitted to the Veterinary Medical Teaching Hospital (VMTH), University of California, Davis, between January 1995 and January 2001 were reviewed retrospectively. All dogs had undergone surgical resection of at least 1 SAC. The diagnosis of SAC was based on myelographic findings and confirmed at surgery. Case histories, neurological examination results, results of CSF analysis, neuroradiological and surgical findings, and outcome were reviewed. All dogs had a CBC, serum biochemistry, urinalysis, abdominal ultrasound, and thoracic radiographs done before anesthesia. In all dogs, plain vertebral column radiographs in both lateral and ventrodorsal projections were done before lumbar CSF collection and myelography. Cerebrospinal fluid analysis was done in 13 dogs. For the myelographic studies, 0.45 ml/kg of contrast medium a was injected into the ventral subarachnoid space with a spinal needle placed through the L4 5 or L5 6 interarcuate space. CT was done in 7 dogs and MR imaging was done in 3 dogs. CT images were obtained immediately after myelography by means of contiguous 1.5-mm transverse images with the use of a CT 9800 system. b All MR images were acquired with a 0.4 Tesla scanner. c Noncontiguous transverse images with a 3.5-mm slice thickness and an interslice gap of mm were generated with T1-weighted (T1W) and T2-weighted (T2W) spin echo pulse sequences. Transverse T1W and T2W and sagittal T1W images were obtained. T1W postcontrast images were obtained after 0.02 mmol/kg IV contrast medium. d In all dogs, SACs were surgically resected after a standard approach to the dorsal or ventral vertebral column. Follow-up neurological examinations were done on 11 dogs 4 to 8 weeks after surgery. Outcome was considered to be fair to good if the neurological signs at the time of recheck examination were improved, even if residual neurological deficits were present. Outcome was considered poor if the neurological status had deteriorated since surgery.

2 Retrospective Analysis of Spinal Cysts 691 Fig 1. Lateral myelogram of a 2-year-old, female, spayed Miniature Poodle with a 5-day progressive T3 L3 myelopathy. A large teardropshaped pooling of contrast medium is noted in the dorsal subarachnoid space at T9 10 suggestive of a SAC. Results Signalment and History Eight of the 14 dogs were Rottweilers and 2 were Pugs. A 2 analysis identified a significantly higher number of Rottweilers affected (P.0001) than expected based on breed numbers admitted to the VMTH. None of the dogs had a history of trauma. All dogs had a history of slowly progressive ataxia and paresis. Four dogs apparently had pain according to the owners. One dog underwent surgery for a Hansen type I intervertebral disc (IVD) extrusion 5 years before the diagnosis of SACs at the same location. One dog had undergone 2 craniotomies for removal of an intracranial intra-arachnoid cyst 4 years before diagnosis of the SAC at C2 3. A hemilaminectomy was done at T13- L1 in this dog for removal of a Hansen type I IVD extrusion 4 months before removal of the SAC, which was found at the time of diagnosis of the IVD extrusion. Results of Neurological Examination All dogs were ataxic and paretic and had deficits in proprioceptive placing in the thoracic limbs, pelvic limbs, or both. Three of the 6 dogs with dorsal SACs at C2 3 had more severe ataxia and paresis in the thoracic limbs than in the pelvic limbs. Four dogs had signs of apparent pain on neurological exam. Clinical signs of neurological dysfunction depended on the location of the SAC. Ten dogs had C1 5 myelopathies, and 4 dogs had T3 L3 myelopathies. CSF Analysis Lumbar CSF analysis was available for review in 13 of 14 dogs. CSF analysis identified mild mononuclear pleocytosis in 7 dogs, with the total nucleated cell count ranging from 1 5 cells/ L in 4 dogs and 5 9 cells/ L in 3 dogs (reference range 5 white blood cells [WBC]/ L). In 5 of these dogs, increased protein concentrations were found, ranging from 41 to 216 mg/dl (reference range 25 mg/ dl). CSF from 6 dogs had an increased protein concentration, ranging from 41 to 127 mg/dl, with a normal number and distribution of nucleated cells. Myelographic Findings The common appearance of SACs on myelography was pooling of contrast medium in the subarachnoid space. In 7 dogs, the filling defect in the subarachnoid space had a teardrop appearance (Fig 1). In 6 dogs, apparent narrowing of the spinal cord at the site of the SAC was present. Two dogs had cysts located ventral to the spinal cord (Fig 2). The remainder of the cysts were located dorsal or dorsolateral to the spinal cord. Multiple, multiple-lobed, or bilobed SACs were present in 5 dogs. SACs were considered multiple if there were 2 or more SACs at different locations or if there were 2 or more cysts at 1 location. Multiple SACs at a single location were suspected to be bilobed if intraluminal connections were identified at surgery. Four dogs had spinal cord compression secondary to IVD disease at locations different from that of the SAC, and 1 dog had spinal cord compression secondary to IVD protrusion at the same level as the SAC. Three of the disc extrusions were classified as Hansen type I, and 2 disc protrusions were classified as Hansen type II. One dog had undergone surgery for an extruded IVD (Hansen type I) at T years before the diagnosis of an SAC at T12 13 and IVD extrusion at T13 L1. The SAC was not observed on the myelogram done 5 years earlier. One dog had myelograms performed 2 and 5 months postoperatively. Two months postoperatively, the ventral SAC, although smaller than on the preoperative myelogram, still was present despite fenestration. The 2 dorsal SACs were unchanged. Five months postoperatively, both the ventral and the dorsal

3 692 Rylander et al SACs were smaller when compared to the previous myelogram (Fig 2). In 1 dog, the diagnosis of an SAC at C2 3 had been made 4 months before SAC resection. At that time, the myelogram disclosed IVD disease with secondary compression of the spinal cord at the level of T13 L1, and surgery was performed to decompress the spinal cord and remove disc material because the dog had clinical signs of a T3 L3 myelopathy. Computed Tomography In 7 dogs, CT was performed after myelography. The results of CT imaging were consistent with the myelographic findings. On transverse images, pooling of contrast medium was observed in the subarachnoid space (Fig 3). One dog had contrast medium in the central canal compatible with syringohydromyelia. Syringohydromyelia was localized between vertebral bodies T12 and L2, and the communication between the central canal and subarachnoid space was in close proximity to the SAC at T13 L1. In 4 dogs, the spinal cord was flattened, irregular, atrophied, or misshapen. MR Imaging One dog underwent MR imaging 9 months and 3 years postoperatively to evaluate the spinal cord in the region of the SAC resection. This dog had 1 of 3 SACs surgically fenestrated at C5 6. No SACs were visible on MR imaging at either 9 months or 3 years postoperatively. Spondylosis was identified at C5 6. One dog had MR imaging at C2 3 before surgery (Fig 4). On sagittal T1W and T2W images, the spinal cord diameter appeared decreased over the caudal body of C2 and cranial aspect of C3. On sagittal T1W images, a discrete hypointense region dorsal to the spinal cord was observed at C2 3 and was contiguous with the subarachnoid space. On T2W images, this region was hyperintense, consistent with a fluid-filled structure. Immediately caudal to this structure, the spinal cord appeared focally widened and then tapered to a more normal appearance. Attenuation of the ventral epidural fat column was identified dorsal to the C2 3 disc space. After IV administration of contrast medium, no contrast-enhancing lesions were identified. The changes described dorsal to C2 3 were compatible with a dorsally positioned SAC and secondary spinal cord compression or atrophy. MR imaging of the brain and cranial cervical spinal cord was done in 1 dog 5 months postoperatively. A cystic lesion was identified within the spinal cord at the level of C2. Surgery A dorsal midline approach to the vertebral column was performed with either a laminectomy, hemilaminectomy, or a combination of both in 13 dogs. A ventral slot was done in 1 dog. The dura mater overlying the SAC was incised and reflected to expose the underlying cyst. In all dogs, an attempt was made to completely resect the cyst lining. The dura mater was not closed. One dog with multiple SACs at a single level underwent Fig 2. Lateral myelographic views of a 3-year-old, male, castrated Rottweiler with a 5-month history of a progressive C1 5 myelopathy. The initial myelogram (A) showed multiple or bilobed SACs at C5 6. The arrow demonstrates the dorsal SAC. The dog underwent a ventral slot at C5 6 with resection of the ventral cyst. On the lateral myelographic views at 2 months (B) and 5 months (C) postoperatively, there is a gradual decrease in the size of both dorsal cysts. Note the postoperative collapse of the C5 6 intervertebral disc space. a ventral slot to fenestrate the ventral cyst. The 2 dorsal cysts were not operated on. In 1 dog with SACs at 2 different locations (C3 4 and T10 11), only the T10 11 SAC was resected. Three of the 5 dogs with concomitant IVD disease and secondary spinal cord compression had decompressive surgeries performed and disc material removed from the vertebral canal at the same time the SAC was removed. One of these dogs had Hansen type II IVD disease at the same intervertebral space as the SAC. The other 2 dogs had Hansen type I disc extrusions at locations remote from the site of the SAC.

4 Retrospective Analysis of Spinal Cysts 693 Fig 3. Transverse CT image at T12 with contrast medium in the subarachnoid space of a 12-year-old, male, intact Dachshund with a 3-month progressive T3 L3 myelopathy and a previous right T13 L1 hemilaminectomy performed 5 years before this study for an IVD herniation. The right lamina and pedicle are missing from the previous surgery, and a SAC (arrow) is identified in the left dorsolateral subarachnoid space. One dog was presented with a history consistent with IVD disease. On myelography, compression of the spinal cord at T13 L1 consistent with IVD disease was observed, and a SAC was present at C2 3. The dog underwent decompressive laminectomy and removal of herniated disc material at T13 L1. The SAC was considered to be an incidental finding. Four months later, the SAC was surgically resected to further decompress the spinal cord and prevent future spinal cord injury. Recovery from surgery in 3 dogs with SACs at C2 3 was complicated by hypoventilation and, in 2 of these dogs, postoperative hemorrhage. One dog had a 2nd myelogram and surgery 24 hours after the 1st surgery to remove an extradural hematoma. This dog required mechanical ventilation for 7 days after the initial surgery. One dog underwent mechanical ventilation from days 2 to 11 after the 1st surgery. A 2nd myelogram and surgery to remove a hematoma were performed 6 days after the initial surgery. Both of these dogs recovered and were able to walk with assistance 17 and 25 days, respectively, after the 1st sur- Fig 4. Magnetic resonance images of a 9-month-old, male, intact Pug with a 7-month history of a progressive C1 5 myelopathy. On the sagittal T1 weighted image (A), there is a hypointense region (arrow) dorsal to the spinal cord at C2 3. On the sagittal T2 weighted image (B), the same region is hyperintense (arrow), suggestive of a fluid-filled structure that is contiguous with the subarachnoid space and of the same signal intensity as CSF on T1 and T2 weighted images. On the transverse T1 weighted image (C) a hypointense structure can be seen dorsal to the spinal cord at C2 3 (arrow). These MR features are suggestive of an SAC.

5 694 Rylander et al gery. One dog had hypoventilation 1 day after surgery and was mechanically ventilated for 2 days. Exploratory surgery 24 hours postoperatively did not identify any hemorrhage around the spinal cord. This dog never became ambulatory again. Histopathologic Evaluation Tissue for histopathologic evaluation was submitted from 10 dogs. In all dogs, fibrosis of the subarachnoid membrane with meningeal cell proliferation was observed. In 6 dogs, the histopathologic findings were suggestive of a subarachnoid cyst. In 5 of these 6 samples, the cysts were lined by meningothelial cells. In the 4 biopsy samples for which a histopathologic diagnosis of SACs could not be made, 1 biopsy sample contained multifocal nodular collagen with cleft formation but no epithelial lining, 2 contained mature fibrous connective tissue in the subarachnoid tissue but no cyst formations, and 1 was histologically normal. Outcome Seven of the 11 dogs available for follow-up examinations after surgery improved in their neurological status. One dog was neurologically normal 5 weeks postoperatively. The other 6 dogs were improved but still had neurological deficits. Residual abnormal neurological findings included ataxia, paresis, and deficits in postural reactions. In the remaining 4 dogs, neurological status had deteriorated. One dog was a Rottweiler with multiple SACs at C5 6. Only the ventral SAC was fenestrated. This dog was minimally ambulatory and had been treated with an antiinflammatory dosage of prednisone (0.3 mg/kg/d) since surgery and was still being treated 3 years post-operatively. One dog was improved neurologically when rechecked 4 weeks postoperatively but subsequently deteriorated and was euthanized by the referring veterinarian 4 weeks later without postmortem examination. This dog had 2 SACs, but only the most caudal SAC was resected. One dog had improved neurologically by 4 weeks postoperatively, but at 9 months postoperatively, the owner reported that the dog was ambulatory only for short periods of time and was recumbent most of the time. This situation improved minimally if the dog was treated with an anti-inflammatory dosage of prednisone (0.4 mg/kg/d). One dog had multiple neurological problems. The dog was minimally ambulatory before surgery for the SAC, never became ambulatory after surgery, and eventually was euthanized 5 months postoperatively. Postmortem Examination Postmortem examinations were performed on 2 dogs. One of the dogs was euthanized for an unrelated cause 2 years postoperatively. Postmortem examination findings included bilateral asymmetrical axonal degeneration and myelin loss in spinal cord segment C2 3 with fibrous adhesion of a fragment of the dorsal lamina to the dura mater and severe focal meningeal fibrosis with multifocal mineralization and adhesion to the dorsal funiculi of the spinal cord. The other dog was euthanized 5 months postoperatively. On postmortem examination, marked dural and meningeal fibrosis were observed with dural-meningeal and meningeal-neuropil adhesion at the C2 spinal cord segment. Multifocal cystic cavitation of the dorsal funiculi and marked locally extensive fibrosis with meningeal adhesion involving the intracranial dura mater also were present. Discussion SACs causing neurological deficits have been reported infrequently in dogs. The terminology is misleading because most of the lesions do not have an epithelial lining. Some authors suggest that SACs be termed cavitations or diverticulae. 6,17 Other terms that have been used include leptomeningeal cyst 1 and meningeal cyst. 5,13 In humans, spinal meningeal cysts are classified into 3 major categories: extradural cysts without spinal nerve root fibers (type I), extradural cysts with spinal nerve root fibers (type II), and intradural cysts (type III). 17 The reported cases of SACs in dogs (including the dogs in this report) most closely resemble type III spinal meningeal cysts of humans. As previously reported, there was no age or gender predilection in the dogs of this report for the occurrence of SACs, and the clinical signs of neurological dysfunction were slowly progressive. Eight of the 39 previously reported cases of SACs in dogs occurred in Rottweilers. 9,16 Interestingly, 8 of the 14 dogs in the present study were Rottweilers. Statistically, a higher number of Rottweilers in this report were affected by SACs than would be expected based on breed numbers admitted to the VMTH (P.0001). Congenital malformation is one of the suspected causes of SACs and a genetic component to the occurrence of SACs in Rottweilers may be present. The occurrence of SACs in related dogs 9,11 further strengthens the possibility of a genetic association. It is also possible that large active dogs with heavy heads may be more prone to SAC formation secondary to trauma, 6,16 but other large breed dogs are not represented in this series. Multiple SACs have been reported previously in only 4 dogs. 9,16 In 1 of these dogs, the cysts were located at 2 different anatomical sites, and in the other 3 dogs, 2 cysts were present at the same location. In this series, all 5 dogs with multiple-lobed or bilobed cysts were Rottweilers. It is possible that connections could exist among multiple cysts in the same anatomic location. Although 1 dog had only 1 of 3 SACs at C5 6 resected, a myelogram 5 months after surgery indicated that all SACs were smaller and the SACs were not present on MR imaging 9 months and 3 years after surgery. A connection may have been present among the 3 SACs and removal of the ventral SAC might have allowed drainage of the 2 other SACs. Resection of 1 SAC also might alter CSF dynamics and possibly relieve outflow obstruction in other SACs that are not connected. Most of the previously reported cysts have been located dorsal to the spinal cord In 1 report, 1 dog had a dorsolaterally located SAC, and 2 dogs with multiple SACs had ventrally located cysts. 16 In this series, 2 dogs had dorsolaterally located, bilobed SACs. Two other dogs, in addition to having dorsolaterally located SACs, also had single ventrally located SACs. Proposed etiologies for SACs in dogs include congenital

6 Retrospective Analysis of Spinal Cysts 695 malformation of the arachnoid membrane, 2 4,6 11,13 trauma, 1,10,18 hemivertebra, 5 scoliosis, 12 arachnoiditis, 1,16 and spinal dysraphism. 3 In humans, SACs are thought to be congenital or to result from trauma, surgery, or arachnoiditis. 19 Intervertebral disc disease as a potential factor in SAC formation has been reported in 2 dogs. 10 One of these dogs had IVD herniation diagnosed 5.5 years before the occurrence of the SAC at the same location, and 1 dog had IVD herniation diagnosed at the same time as the SAC at the same location. In this study, 1 dog had surgery for IVD herniation 5 years before the occurrence of SACs at the same site. The SAC was not present at the time IVD disease was diagnosed; therefore, it was unlikely to be congenital in origin. The proximity of the SAC to the previous IVD herniation in this dog and the occurrence of spinal cord compression secondary to IVD disease at the same level as the SAC in another dog in this study might suggest that trauma to the meninges contributed to cyst formation. The abnormal CSF findings in dogs of this study differ from findings in other studies, in which CSF was reported to be normal In previous studies, spinal fluid was collected from the cerebellomedullary cistern cranial to the SAC. The likelihood of detecting CSF abnormalities increases when spinal fluid is collected caudal to the lesion. 20 The occurrence of mononuclear pleocytosis and increased protein concentration in the CSF in dogs of this report could reflect chronic irritation of the meninges associated with SACs. It is also possible that existing arachnoiditis induced formation of the SAC. When different imaging techniques were compared, it was found that myelography enabled visualization of all SACs, but CT better defined any lateralization of the cysts. 10 MR imaging was helpful in assessing spinal cord parenchyma and in detecting syringohydromyelia. However, because filling of the SAC depends on patient positioning 18,19 and the location of the SAC, SACs might be diagnosed more readily with myelography than with other imaging techniques in which the patient is imaged in only 1 position. In humans, CT myelography could be the imaging modality of choice because it depicts the anatomy of the dural defects in more detail than MR imaging. 19 Other studies have concluded that MR imaging is the diagnostic procedure best suited for the detection of SACs because it reveals the exact location, extent, and relationship of the lesion to the spinal cord. 21 In addition, spinal cord atrophy, which can be used to predict neurological outcome, might be better visualized by MR imaging. 21 All SACs in dogs in this study were diagnosed based on myelographic findings. CT was useful to confirm the myelographic findings, define the exact location of the cysts, and plan surgery. A diagnosis of syringohydromyelia associated with SAC was made based on CT in 1 dog. MR imaging helped exclude involvement of the spinal cord parenchyma in 1 dog. Insufficient MR or CT imaging data was available to draw any conclusions about their usefulness in treatment and prognosis. Of the 39 previously reported cases of SACs in dogs, 30 underwent surgery. Of the 19 dogs that were followed up postoperatively, 13 dogs were neurologically normal, 4 dogs were improved but had neurological deficits, and 2 dogs had poor outcome and were euthanized. One of the 39 dogs was treated with an anti-inflammatory dosage of prednisone and improved neurologically. Eight dogs were euthanized after the myelogram. Only 5 dogs were followed more than 10 months postoperatively. In 1 study 9 with follow up ranging from 10 months to 4 years postoperatively, 3 of 4 dogs with SACs had a good outcome after durectomy. The 4th dog had 3 separate surgeries for 2 different SACs and recurrence of one of the SACs. This dog was euthanized because of progression of signs. In this series, 1 dog was neurologically normal 4 weeks postoperatively, 6 dogs improved but had residual neurological deficits, and 4 dogs had poor outcomes. One disadvantage of this study is lack of long-term follow-up in most of the dogs. One dog with multiple SACs has been followed for more than 3 years. This dog has continued to slowly deteriorate despite resolution of the SAC on myelography and MR imaging. This dog had multiple SACs. Three of the 6 dogs with SACs at C2 3 reported here experienced a sudden postoperative deterioration in neurological signs resulting in hypoventilation. In 2 dogs, a 2nd surgery to remove a hematoma within the vertebral canal at the site of surgery was done. In a report of ventilatory failure after cervical spinal surgeries, 22 it was suggested that respiratory problems were more common in dogs undergoing dorsal decompressive surgery from the 2nd to 4th cervical vertebrae. The histopathologic findings in the 10 SAC biopsy specimens were similar to those of previous reports. 4,5,7 13 An epithelial lining was not present in any of the 10 submitted samples, and a cleft in the extradural collagen was present in 1 dog. An epithelial lining is not a criterion for diagnosis of SACs, and the finding of cleft formation is similar to 1 previously reported case of SAC. 10 The normal histopathologic appearance of the meninges in 1 specimen underscores the difficulty of resecting the cyst lining and obtaining a representative biopsy. The findings in the 2 dogs that were necropsied suggest that, despite surgery, residual chronic spinal cord pathology can exist and progress. The cystic lesion seen within the spinal cord on MR imaging in 1 dog 5 months postoperatively was confirmed at postmortem examination and was suspected to be a sequela to malacia. Diagnosis and surgical intervention in the disease process before chronic myelopathic changes have developed might influence outcome. SACs should be included in the differential diagnosis in dogs of any age and breed with signs of myelopathy. A higher incidence of SACs occurs in Rottweilers, SACs can be multiple-lobed or bilobed, and they can be associated with IVD disease. Myelography effectively localizes SACs in dogs, and CT and MR imaging might provide useful additional information. Surgical resection results in a reasonable outcome in most affected dogs, but neurological deficits can persist. Footnotes a Iopamidole 41%, Isovue 200, Bracco Diagnostics Inc, Princeton, NJ b CT 9800 system, GE Medical Systems, Milwaukee, WI c 0.4 Tesla scanner, Resonex 5000, Resonex Inc, Sunnyvale, CA

7 696 Rylander et al d Gadolinium-diethylenetriaminepenta-acetic acid (DTPA) dimeglumine, Magnevist, Berlex Laboratories Inc, Wayne, NJ Acknowledgments We thank Jacqueline L. Grandy, DVM, for reviewing the manuscript, John Doval for medical illustrations, and Philip Kass, DVM, for the statistical analysis. References 1. Gage ED, Hoerlein BF, Bartels JE. Spinal cord compression resulting from a leptomeningeal cyst in the dog. J Am Vet Med Assoc 1968;152: McKee WM, Renwick PW. Marsupialisation of an arachnoid cyst in a dog. J Small Anim Pract 1994;35: Dyce J, Herrtage ME, Houlton JEF, Palmer AC. Canine spinal arachnoid cysts. J Small Anim Pract 1991;32: Bentley JF, Simpson ST, Hathcock JT. Spinal arachnoid cyst in a dog. J Am Anim Hosp Assoc 1991;27: Parker AJ, Smith CW. Meningeal cyst in a dog. J Am Anim Hosp Assoc 1974;10: Parker AJ, Adams WM, Zackery JF. Spinal arachnoid cysts in the dog. J Am Anim Hosp Assoc 1983;19: Cambridge AJ, Bagley RS, Britt LG, Silver GM. Radiographic diagnosis: Arachnoid cyst in a dog. Vet Radiol Ultrasound 1997;38: Hashizume CT. Cervical spinal arachnoid cyst in a dog. Can Vet J 2000;41: Frykman OF. Spinal arachnoid cyst in four dogs: Diagnosis, surgical treatment and follow-up results. J Small Anim Pract 1999;40: Galloway AM, Curtis NC, Sommerlad SF, Watt PR. Correlative imaging findings in seven dogs and one cat with spinal arachnoid cysts. Vet Radiol Ultrasound 1999;40: Ness MG. Spinal arachnoid cysts in two Shih Tzu littermates. Vet Rec 1998;142: Bagley RS, Silver GM, Seguin B, et al. Scoliosis and associated cystic spinal cord lesion in a dog. J Am Vet Med Assoc 1997;211: Hardie RJ, Linn KA, Rendano VT. Spinal meningeal cyst in a dog: A case report and literature review. J Am Anim Hosp Assoc 1996;32: Vignoli M, Rossi F, Sarli G. Spinal subarachnoid cyst in a cat. Vet Radiol Ultrasound 1999;40: Shamir MH, Shahar R, Aizenberg I. Subarachnoid cyst in a cat. J Am Anim Hosp Assoc 1997;33: Gnirs K, Ruel Y, Blot S, et al. Which lesions characterize spinal sub-arachnoid cysts: Study in 11 dogs. Vet Radiol Ultrasound 2001; 42: Nabors MW, Pait TG, Byrd EB, et al. Updated assessment and current classification of spinal meningeal cysts. J Neurosurg 1988;68: Kriss TC, Kriss VM. Symptomatic spinal intradural arachnoid cyst development after lumbar myelography. Case report and review of the literature. Spine 1997;22: Myles LM, Gupta N, Armstrong D, Rutka JT. Multiple extradural arachnoid cysts as a cause of spinal cord compression in a child. Case report. J Neurosurg 1999;91(1 Suppl): Thomson CE, Kornegay JN, Stevens JB. Analysis of cerebrospinal fluid from the cerebellomedullary and lumbar cisterns of dogs with focal neurologic disease: 145 cases ( ). J Am Vet Med Assoc 1990,196: Krings T, Lukas R, Reul J, et al. Diagnostic and therapeutic management of spinal arachnoid cysts. Acta Neurochir (Wien) 2001; 143: Beal MW, Paglia DT, Griffin GM, et al. Ventilatory failure, ventilatory management and outcome in dogs with cervical spinal disorders; 14 cases ( ). J Am Vet Med Assoc 2001;218: