Spinal Arachnoid Cysts in 17 Dogs

Statistical Analysis

Clinical trends in presentation were identified from the records, and because cysts tended to occur in two locations (i.e., cervical and thoracolumbar spine), dogs were grouped according to cyst location (cervical and thoracolumbar). The body weights and ages of dogs were compared between the groups using the Wilcoxon’s rank sum test, with significance set at P<0.05. The presence of incontinence (fecal, urinary, or both) was compared between the groups using the Fisher’s exact test, with significance set at P<0.05.

Short-term outcome (<1 month postoperatively) was determined in 15 dogs that underwent surgery by repeat physical examination by the authors. Long-term outcome (>1 year) was determined in 12 of the 15 dogs that underwent surgery by repeat physical examinations by the authors (n=8), or by telephone communications with the owner together with evaluation of the medical record (n=4). Two dogs were excluded because they had surgery within the last 6 months. One additional dog (case no. 12) was excluded from this analysis because she was euthanized as a result of another disease 2 months postoperatively. Case no. 10 was included in this analysis, despite being euthanized at 10 months postoperatively, because his death was a direct consequence of the arachnoid cyst. Outcome was deemed unsuccessful if there was deterioration in the neurological status or if euthanasia was performed as a result of neurological disease. As most dogs were walking at the time of diagnosis, outcome was deemed successful if there was no further deterioration in neurological status or if there was an improvement in neurological status. Factors (i.e., location of cyst, age at time of diagnosis, duration of clinical signs, and surgical technique) that might be related to long-term outcome following surgery were identified in these 12 dogs. A Fisher’s exact test was used to compare these factors to long-term outcome to establish whether they were significant predictors of outcome. Significance was set at P<0.05.

Cervical Arachnoid Cysts

Arachnoid cysts in the cervical region were diagnosed in seven dogs. Details of the findings in these dogs are summarized in Table 1. Six dogs were large breed (>20 kg), with a mean weight of 33 kg (range, 3 to 67 kg) and a mean age of 31 months (range, 7 months to 7 years). All dogs presented with tetraparesis and ataxia characterized by a hypermetric gait in the thoracic limbs. Cervical pain was not noted in any of the dogs, and urinary and fecal incontinence was present in only one dog. Urinary incontinence was characterized by an inability to generate an uninterrupted flow of urine and continued dribbling after urination, consistent with urethral dyssynergia. The perineal reflex was normal in the dog with fecal incontinence, and the condition improved with a high-fiber diet. There was no history of trauma in any dog. Duration of clinical signs prior to presentation ranged from 2 weeks to 3 years (mean, 7.5 months), and the signs were progressive in all dogs.

Minimum database, cisternal CSF analysis, and survey radiographs were unremarkable except for identifying a urinary tract infection and diskospondylitis of the third and fourth thoracic (T₃–T₄) vertebrae in case no. 7. Myelography revealed a bulbous expansion of the subarachnoid space consistent with a diagnosis of arachnoid cyst [Figure 1] in six of seven dogs. Case no. 5 had an intradural filling defect on myelography [Figure 2]; the diagnosis of arachnoid cyst was made from CT images in which contrast was present in the dilated subarachnoid space. In case no. 6, there was a typical bulbous expansion of the subarachnoid cyst at C₁ and the second cervical (C₂) vertebrae, and contrast was also present within the spinal cord parenchyma ventral to the cyst [Figure 3A]. Magnetic resonance images confirmed a syrinx in the subjacent spinal cord [Figures 3B, 3C]. In case no. 7, there was an extradural compressive lesion immediately ventral to the cyst at the sixth and seventh cervical (C₆–C₇) vertebrae that was consistent with a diagnosis of caudal cervical spondylomyelopathy (CCSM) [Figure 4]. Computed tomography scans highlighted the lateralization of cysts, but otherwise contributed little additional information in the three dogs in which the cyst was visible on myelography. The cystwas located at the C₁ to C₃ vertebral level in four dogs and extended from C₂ to the fifth cervical (C₅) vertebrae in one dog. Two dogs (both Great Danes) had a cyst dorsal to the C₇ vertebrae; in case no. 7, the cyst was directly dorsal to an extradural compressive lesion at the C₆ to C₇ vertebrae.

All dogs with cervical arachnoid cysts underwent surgical decompression. In six dogs, the cyst was approached by a dorsal laminectomy; case no. 2 had a right-sided cyst that was approached via a right hemilaminectomy. In all dogs, a large volume of CSF was released when a durotomy was performed over the cyst. The cysts were marsupialized in five dogs and fenestrated in two dogs. Case no. 6 also had a shunt placed to divert fluid within the syrinx to the cisterna magna. This was accomplished by placing a small section of silastic tubing from the syrinx to the cisterna magna and suturing it to the dura mater via 9-0 nylon. Further details of this shunting procedure were not available from the medical record. Intraoperative ultrasonography was used to delineate the boundaries of the syrinx for shunt placement. The excised meninges/cyst wall was submitted for histopathology in two dogs and was described as normal meninges in each instance. The short-term surgical outcome was good in all dogs. The long-term surgical outcome was good in four of six dogs with sufficient follow-up time. Two dogs improved initially following surgery but developed a recurrence of neurological signs at 18 and 26 months postoperatively. Further diagnostics or necropsies were not permitted in either case. Case no. 7 was initially treated with antibiotics alone because of concurrent diskospondylitis and CCSM. The dog’s ataxia progressed over 16 months, at which point he became very exercise intolerant. Surgery was recently performed, and the dog has improved in the 2 months since surgery.

Thoracolumbar Cysts

Ten dogs were diagnosed with thoracolumbar arachnoid cysts [Table 2]. In contrast to the dogs with cervical arachnoid cysts, seven dogs were small breeds weighing <20 kg, and only three were large breeds (overall mean weight, 20.2 kg). Only two dogs were <1 year of age; the remainder ranged from 3 to 11 years of age (mean, 6.2 years). Duration of clinical signs prior to presentation ranged from 3 weeks to 4 years (mean, 11.3 months). All dogs presented with upper motor neuron (UMN) pelvic limb ataxia and paresis, and nine of 10 dogs had a progressive clinical course. The pelvic limb gait was hypermetric in four dogs, and three dogs had a long history of “bunny hopping” when walking fast or running. Fecal or urinary incontinence, or both, was present in six dogs. As in dogs with cervical arachnoid cysts, urinary incontinence was typically an inability to generate an uninterrupted flow of urine and continued dribbling after urination. In all dogs, the perineal reflex was normal. Only one dog experienced pain on palpation of the spine.

The minimum database, CSF analysis, and survey radiographs were unremarkable in all dogs. Urinary tract infections were present in three dogs at presentation or in the follow-up period. Myelography revealed a bulbous expansion of the subarachnoid space and was diagnostic of an arachnoid cyst in eight of 10 dogs [Figure 5]. In case nos. 12 and 13, intradural filling defects were identified on the myelograms. The diagnosis of arachnoid cyst was made from MR images in case no. 12 and at surgery in case no. 13. Computed tomography was performed in eight dogs and added little to the clinical picture except in case no. 10, where severe attenuation of the spinal cord was noted. In eight dogs, the cyst was located dorsally over the caudal thoracic (T₁₁–T₁₃) vertebrae, and in two of these dogs, the cyst spanned more than two vertebrae. Cysts located at different sites in the lumbar spine were identified in the remaining two dogs, and in one of these dogs, the cyst was lateralized to the right.

Case no. 15 was hit by a car and suffered a fracture/luxation of the 12th thoracic (T₁₂) and T₁₃ vertebrae and a fracture of the sixth lumbar (L₆) vertebrae 4 years previously. He was paraplegic at the time of that injury, but he regained pelvic-limb function following surgical stabilization of the T₁₂–T₁₃ fracture and decompression of the L₆ fracture. He also has a long history of effusions and pain in both stifles. Three years after surgery, his pelvic-limb gait worsened and subsequent myelography revealed an arachnoid cyst at the thoracic surgical site [Figure 6]. There was no cyst present on the previous myelogram. At this time it was unclear whether his gait had deteriorated because of neurological disease or degenerative joint disease of his stifles. Conscious proprioceptive deficits were evident in both pelvic limbs, but the patient had never regained normal conscious proprioception following the initial injury. Severe pain was elicited upon manipulation of both stifles, with crepitus and joint effusions, and a presumptive diagnosis of bilateral partial cranial cruciate ligament tears in both stifles was made. The dog has been treated conservatively (i.e., restricted activity), and his clinical signs have remained static in the 21 months since diagnosis.

The spinal cord was surgically decompressed in eight dogs with thoracolumbar cysts. In seven dogs, the spinal cord and cysts were exposed by a dorsal laminectomy, and in one dog by a hemilaminectomy. Fenestration of the cyst was performed in three dogs, the cyst was marsupialized in four dogs, and it was removed intact in one dog. The complete resection was accomplished by gently stroking the dorsal midline of the dura mater with a no. 11 scalpel blade until the cyst-like structure emerged through the defect in its entirety. It was not submitted for histopathological evaluation. Intraoperative abnormalities associated with the cyst were noted in two dogs; in case no. 9, there were thickened trabeculae between the arachnoid and pia mater, and in case no. 12, the dorsal half of the spinal cord appeared to bifurcate directly ventral to the cyst. Histopathology performed on excised cyst wall in four dogs revealed normal meninges.

A good outcome was achieved within 1 month in seven of the eight dogs treated surgically. Case no. 10 showed no improvement after surgery and was euthanized 8 months postoperatively, at which time he was paraplegic. Two additional dogs worsened within 14 months of surgery, and only five dogs had a good long-term outcome. Case no. 12 improved initially but developed tetraparesis and was euthanized 2 months postoperatively. Globoid cell leukodystrophy and spinal dysraphism (third to seventh lumbar [L_3–L₇] vertebrae) were diagnosed at necropsy; there was no evidence of cyst recurrence. Case no. 13 improved initially but developed severe paresis 14 months postoperatively and was euthanized. Case no. 14 was stable initially but was thought to have developed urinary and fecal incontinence 2 months postoperatively. Myelography and CT were repeated, and there was no evidence of cyst recurrence [Figures 7A, 7B]. However, the spinal cord was mildly compressed laterally at the surgical site [Figure 7C]. Further questioning of the owner revealed that the dog was incontinent before surgery and she had mistakenly blamed a puppy in the household. The dog was judged to have a successful outcome because no further deterioration in gait or continence occurred postoperatively. Case no. 15 has not undergone surgery and has been stable in the 21 months since diagnosis. His pelvic-limb weakness appears to worsen when he exercises; however, this is ascribed to his stifle disease. Case no. 16 was euthanized immediately after diagnosis, and necropsy confirmed the presence of a dilated subarachnoid space dorsal to the spinal cord from T₁₂ to the second lumbar (L₂) vertebrae. The subjacent spinal cord had moderate to marked gray and white matter atrophy and axonal degeneration. Case no. 17 recently had surgery and has improved voluntary motor ability, but long-term follow-up is not yet available.

Clinical Trends and Outcome

Dogs with cervical cysts were significantly larger and younger than dogs with thoracolumbar cysts (P=0.009 and P=0.037, respectively). Incontinence was commonly present in dogs with thoracolumbar cysts, but there was only a trend toward significance (P=0.08) [Table 3]. In the short term, 14 of 15 (93%) dogs had a good outcome postoperatively. However, in the long term, only eight of 12 (67%) dogs maintained a good outcome postoperatively. Case no. 15 had a good long-term outcome with conservative treatment but was excluded from long-term analysis because he did not undergo surgery. Although no significant predictors of outcome were identified, several factors were associated with a good long-term outcome. The age at diagnosis, the duration of clinical signs prior to diagnosis, and the surgical technique all had a trend toward significance. Successful long-term outcomes were seen in all of the five dogs that were <3 years of age, compared with three of seven dogs that were ≥3 years of age (P=0.08). Successful long-term outcomes were seen in seven of eight dogs with a duration of signs of ≤4 months, compared to only one of four dogs with clinical signs present for >4 months (P=0.07). Six of seven dogs treated by marsupializing the cyst had successful outcomes, compared with two of five dogs treated by cyst fenestration (P=0.15).

Etiology

Spinal arachnoid cysts are rare in veterinary medicine and have been reported in only 28 dogs and three cats since 1968.^1–15 Many of these cases have been reported in the last 5 years, suggesting that spinal arachnoid cysts are being diagnosed with increasing frequency. This may reflect improvements in imaging of the central nervous system, although the majority of cases are diagnosed by myelography alone. As most previous reports of canine spinal arachnoid cysts describe single or low numbers of cases, the aims of this paper were to identify common traits and evaluate long-term outcome in a larger group of dogs.

The pathophysiology of arachnoid cyst formation remains unknown, but there is evidence that such cysts can be congenital or acquired.^1–18 As no single theory can account for the formation of all cysts, there are probably several different etiologies. It is plausible that spinal arachnoid cysts in young dogs represent congenital lesions. Most theories on congenital cyst formation postulate that cysts expand with time because of one-way valves that intermittently allow accumulation of CSF within the cyst as CSF pressure fluctuates.³¹⁶ Thus, clinical signs are only seen once a cyst is large enough to cause significant spinal cord compression, and arachnoid cysts are not always clinically significant.³ As the cranial cervical and thoracolumbar spinal segments are very mobile, it is conceivable that pressure changes exist within the subarachnoid space of these two areas and predispose them to cyst formation.

The dogs described in this paper could be grouped based on cyst location. Cysts most commonly occurred on the dorsal midline over C₁ to C₃ or T₁₁ to T₁₃. Dogs with cranial cervical cysts tended to be large-breed, young dogs, whereas those with caudal thoracic cysts were typically small-breed dogs and were often diagnosed later in life. These findings are consistent with those already reported in the veterinary literature^1–15 [Table 4] and suggest that there is something specific to these two sites that predisposes to cyst formation. Only four dogs in this study had cysts at locations other than C₁ to C₃ or T₁₁ to T₁₃, and two of thesedogs had additional spinal cord diseases that could have contributed to cyst formation at these sites.

One dog in the current study developed a spinal arachnoid cyst 4 years after spinal cord trauma due to a vertebral fracture. A dog and a cat have been reported to develop arachnoid cysts at the site of previous intervertebral disk herniations.⁴ None of these animals had evidence of arachnoid cyst formation on myelography at the time of spinal cord injury, suggesting that the cysts were acquired secondary to the traumatisc injury. The development of arachnoid cysts following spinal cord trauma has also been reported in humans.¹⁹ It is postulated that trauma damages the arachnoid mater, resulting in scarring and, hence, cyst formation. Another dog in the current study had a large arachnoid cyst dorsal to a ventral, extradural compressive lesion. Again, it seems likely that this cyst was acquired as a result of the coexisting compressive spinal cord lesion, although in this dog there was no previous myelogram to prove that the cyst was acquired. Arachnoiditis is also proposed as a cause of acquired cysts.⁷ However, lack of inflammation in the meningeal biopsies or in CSF samples from the dogs in this and other studies^1–15 makes this etiology less likely.

Clinical Findings

Typical neurological signs exhibited by the dogs in this study included progressive paresis with a hypermetric gait, fecal or urinary incontinence, or both, and no evidence of spinal hyperesthesia. Hypermetria was a common finding in this study, in contrast to previous reports [Table 4], but the authors have no clear explanation for this discrepancy. The spinocerebellar tracts lie along the dorsolateral aspects of the spinal cord²⁰ and as such, may be compressed by the cysts dorsal to them, resulting in hypermetria. Urinary and fecal incontinence were also consistent findings (especially in dogs with thoracolumbar cysts) and have been noted in other dogs with spinal arachnoid cysts at both the thoracic and cervical locations.²³⁵⁷¹³ Incontinence is also a common finding in humans with type III arachnoid cysts.¹⁶¹⁷ Loss of continence in ambulatory dogs with UMN signs (to the pelvic or all four limbs) is unusual in canine spinal cord disease, and it may again reflect the dorsal location of the arachnoid cysts. The presence of urinary or fecal incontinence in an ambulatory animal with UMN signs and a hypermetric gait should heighten the clinician’s index of suspicion for the presence of a spinal arachnoid cyst.

Diagnostic Findings

Spinal arachnoid cysts were diagnosed on myelography in the majority of dogs in this study. In three dogs, the cyst appeared as a filling defect, suggesting an intradural lesion that did not communicate with the subarachnoid space. This is a common finding in human patients and is thought to be a result of the presence of one-way valves that open only periodically with changes in CSF pressure.¹⁶ Another possible explanation is that the opening into the cyst in these dogs was located in such a position that contrast could only flow into the cyst from one direction. Computed tomography facilitated diagnosis of one of the cysts that appeared as a filling defect on myelography, probably because less contrast media is needed for visualization on CT.

In humans with cervical spondylotic myelopathy, the degree of spinal cord attenuation is related to outcome.²¹ It is possible that this is also true in other chronic compressive spinal cord diseases,²² and CT images provide an accurate means of assessing spinal cord attenuation.⁴²¹ The number of dogs that had a CT myelogram performed in the current study was not adequate to evaluate the relationship between spinal cord attenuation and outcome. However, it was notable that the dog (case no. 10) with the most severely attenuated spinal cord had a poor outcome after surgery. Magnetic resonance imaging was useful in diagnosing a noncommunicating cyst (case no. 12) and in confirming a syrinx in case no. 6, as was reported in another study.⁴

Treatment and Outcome

Treatment of spinal arachnoid cysts in dogs has historically involved surgical decompression of the spinal cord with fenestration to prevent recurrence. In previous reports, surgical decompression produced a good outcome in most dogs, although signs did recur in some [Table 4].^1–15 However, close evaluation of the literature revealed that only three dogs were followed for more than 1 year. As there is a tendency for single case reports to describe animals with a successful outcome, the authors wanted to evaluate the long-term outcome in all dogs with spinal arachnoid cysts treated surgically at NCSU-VTH. Although 14 of 15 dogs improved in the short term, five of these dogs did not have a good long-term outcome. Of these five dogs, one was euthanized because it had globoid cell leukodystrophy; the worsening of neurological signs was not associated with cyst recurrence. Myelography and necropsy were not performed in the remaining four dogs, and it is not known whether the cyst reformed or the signs were the result of another disease process. In three of the five dogs that did not have a good outcome, recurrence of neurological signs developed from 14 to 26 months after surgery.

There is little information on medical management of dogs with spinal arachnoid cysts. There is one report of a dog with a cervical arachnoid cyst that responded well to treatment with a tapering dose of prednisone.⁷ One dog in this study was not treated surgically because of concurrent disease that may be contributing to his gait deficits. He has remained stable in the 21 months since diagnosis, although he does have neurological deficits. Because the clinical significance of the cysts is not known, it is difficult to draw conclusions about conservative management from these dogs. In nearly all dogs (16 of 17), the clinical signs were progressive, and surgical treatment appeared to alter the course of the disease.

Several factors appeared to affect outcome in the dogs of this study. Although they were not statistically significant, they approached significance and may have achieved this if more cases were available. Firstly, six of seven animals <3 years of age had improved postoperative neurological function and successful outcomes. Case no. 4 did not have a successful outcome, but it may have had inadequate cyst drainage because the cyst margins extended beyond the laminectomy site. Secondly, duration of clinical signs for ≤4 months was associated with a better outcome. This may reflect increased axonal loss with increased duration of compression. Lastly, there was a trend toward a better long-term outcome if marsupialization was the surgical technique. Recurrence of arachnoid cysts has been reported in humans¹⁸ and dogs⁶¹³ when fenestration was performed, and thus marsupialization is recommended if complete resection is not possible. A previous veterinary report advocating marsupialization⁶ described suturing the meninges to the endosteum of the vertebral canal, which is technically difficult in the authors’ opinion. The authors have found the meninges can be easily marsupialized to the surrounding paraspinal fascia, provided that the sutures are not placed too tightly, which may result in tension on the spinal cord. Although there was a trend toward significance with surgical technique, confounding factors may be responsible for the poor outcome of three fenestration cases. The neurological deterioration in these cases may be due to other disease (e.g., syrinx in case no. 6) or the presence of spinal cord atrophy (e.g., spinal cord attenuation on CT of case no. 10). Repeat diagnostic tests were only performed in one case of suspected recurrence, and thus other diseases may have accounted for the deterioration. Nonetheless, marsupialization is recommended because of the apparent trend toward a better outcome, the ease of the procedure, and the recommendation that it be performed in humans with the condition.

The presence of concurrent spinal cord disease may affect outcome, but the small number of cases with concurrent disease precluded statistical analysis. In the current study, arachnoid cysts were diagnosed in four dogs (case nos. 6, 7, 12, 15) that had concurrent spinal cord diseases. Case no. 6 had a syrinx ventral to the cyst identified on MR images. Case no. 7 had diskospondylitis affecting T₃ and T₄ and extradural compression of the spinal cord ventral to the cyst at C₆ to C₇. Case no. 12 had spinal dysraphism ventral to the cyst at L₃ to L₇ and globoid cell leukodystrophy. Case no. 15 had a spinal fracture at the level of the cyst. Three of these dogs (case nos. 6, 7, 12) were treated surgically and had good short-term outcomes. Long-term outcome is not available yet for case no. 7, but outcomes were poor for case nos. 6 and 12. Case no. 15 was managed conservatively, because it was unclear whether the cyst was causing the clinical signs.