Thoracolumbar spinal arachnoid diverticula in 5 pug dogs

Abstract

Clinical features, myelography, and computed tomography imaging findings as well as neurological outcome with and without surgery in 5 pug dogs with thoracolumbar arachnoid diverticula are described. Short-term prognosis after surgical therapy may not be as good as reported for other canine breeds, since immediate postsurgical deterioration is possible. Improvement of neurological deficits beyond the presurgical status may take several months.

Spinal subarachnoid cysts in dogs causing neurological deficits have been reported previously (1–19). There is an ongoing debate regarding the correct terminology for those subarachnoid dilatations. The terms pseudocysts, cavitations, or diverticulae have been advocated in veterinary medicine since those structures are lacking true epithelial linings (3,18,19). The term arachnoid diverticulum will be used in this report, except when other authors using a different term are referenced.

In human medicine, 3 types of spinal meningeal cysts have been described: extradural cysts without nerve root fibers (type I), extradural cysts with nerve root fibers (type II), and intradural cysts (type III) (20). All cases reported in dogs most closely resemble type III spinal meningeal cysts (15). A total of 88 canine cases involving various breeds have been reported; most of these involve large breed dogs with predominantly cervical subarachnoid cysts (15–18). Specific features of thoracolumbar arachnoid diverticula in 5 pug dogs are reported here.

Case descriptions

Case 1

A 7.5-year-old male neutered pug dog was presented with a 2-week history of difficulty walking on the pelvic limbs. The dog had been diagnosed with an infarction of the middle cerebellar artery based on MRI 3 years prior to the current presentation. Neurological examination revealed ambulatory paraparesis and mild pelvic limb ataxia. The response to the paw replacement test was severely reduced in the right pelvic limb and completely lost in the left pelvic limb, whereas segmental spinal reflexes were normal in both pelvic limbs. There was no pain on spinal palpation. Signs were consistent with a T3–L3 spinal cord lesion. Myelography was performed using a lumbar injection of 100 mg/kg body weight (BW) Iopamidol (Solutrast 200M; Bracco Imaging, Deutschland GmbH, Konstanz, Germany) at L5–6. Images were generally taken in laterolateral projection only, since once the lesion was identified ventrodorsal views were renounced in favor of computed tomography (CT). There was an extension of the dorsal subarachnoid space with a drop-shaped appearance starting at the cranial border of T11 and slowly fading towards the caudal aspect of the T8 vertebra resulting in the diagnosis of an arachnoid diverticulum (Figure 1). Spinal cord height at the maximal diverticulum extension was reduced by 36% compared to the corresponding localization 1 vertebra caudally. This measurement was taken to estimate the size of the diverticulum and associated reduction of the spinal cord diameter knowing that a summation image such as a radiograph may not accurately reflect the size of the diverticulum due to superimposition of spinal cord and diverticulum. In addition, there was a mild disc protrusion at the intervertebral disc (IVD) space T10–11. Computed tomography (CT) was performed immediately following myelography to evaluate full diverticulum extension with regard to the spinal circumference for surgical planning. The following CT settings were used: slice thickness: 1 mm, pitch: 0.6, collimation: 6 × 0.75, 140 kVp, 250 mAs/slice (Philips brilliance CT 6 slice scanner; Philips GmbH, Hamburg, Germany). Image reconstruction was done using an image-matrix of 512 × 512. Dorsal diverticulum localization was confirmed where it had already been identified on myelography. It was more obvious on sagittally reconstructed images than on transverse images, since the later required direct comparison with transverse images of a normal area.

Figure 1.

Lateral myelography of the thoracic spine of a 7.5-year-old pug dog. Note the drop-shaped dorsal arachnoid diverticulum at T11 (arrow).

Surgery was performed using a dorsal laminectomy from T9 to T12. Dura mater overlying the lesion identified by bluish discoloration was partially resected, creating an opening of about 15 × 5 mm. The defect was covered by a thin layer of a moistened gelatin sponge (Gelaspon; Bausch and Lomb GmbH, Berlin, Germany). Closure was done in a routine fashion.

Ambulatory paraparesis and pelvic limb ataxia worsened to paraplegia with intact deep pain sensation immediately after surgery. The dog slowly recovered neurological function reaching the presurgical condition 3 mo after surgery. At an 8-month reevaluation, the dog had further improved and was ambulatory with a mild paraparesis and slightly reduced response in paw replacement tests in both pelvic limbs.

Case 2

A 5.5-year-old male intact pug dog was presented with a 4-month history of scuffing both pelvic limbs and spontaneous knuckling over, causing skin lesions dorsally on the digits of both pelvic limbs. Clinical signs did not improve following nonsteroidal anti-inflammatory medication. Neurological examination revealed an ambulatory paraparesis and a mild pelvic limb ataxia. The response to the paw replacement test was reduced in the right and lost in the left pelvic limb. Patellar tendon reflexes were bilaterally increased whereas the other segmental spinal cord reflexes were normal. There was no pain on spinal palpation. Signs were consistent with a T3–L3 spinal cord lesion. Lumbar myelography revealed an attenuation of the dorsal contrast column over the IVD space T7–8 and a suspected dilated subarachnoid space from the cranial aspect of the vertebral body T8 to the caudal aspect of the vertebral body T11. A hemivertebra at T6 and spondylosis deformans at T9–10 were the only other pathological findings within 2 vertebral bodies of the lesion. A definite diagnosis, however, could not be made from this laterolateral view only. Computed tomography revealed a mainly left lateralized and slightly dorsal arachnoid diverticulum from the caudal aspect of the T8 vertebra to the caudal aspect of the T10 vertebra, reducing the normal spinal cord diameter at the maximum extension of the diverticulum by 62.5% (Figure 2). To evaluate the extent of intramedullary changes spinal magnetic resonance imaging (MRI) of the lesion identified on CT-myelography was performed, using a 0.5 Tesla magnet (Gyroscan NT Compact Plus; Philips Medical Systems, Eindhoven, The Netherlands). Sagittal T2-weighted images were obtained using the following MRI settings: TR 2250 ms, TE 100 ms, slice thickness 3 mm, interslice gap 0.3 mm. Intramedullary signal hyperintensity was seen from T7 to T10. Differentiation between intramedullary edema, dilated arachnoid space, and epidural fat was not possible on this sagittal view.

Figure 2.

Transverse myelogram computed tomography of a 5.5-year-old pug dog with a left-dorsal arachnoid diverticulum at T9. The hyperattenuating diverticulum is occupying the majority of the transected spinal canal and there is only a thin rim of spinal cord tissue left (arrows).

A left-sided hemilaminectomy from T8 to T10 combined with a durectomy reaching from slightly cranial of the cyst to the cranial aspect of T10 was performed. The defect was covered by a thin layer of a moistened gelatin sponge. Closure was done in a routine fashion. Postoperatively the dog was nonambulatory and paraparetic with minimal voluntary motor function in the left pelvic limb. On a 5-month follow-up examination the dog showed an ambulatory right pelvic limb monoparesis. The left pelvic limb had completely recovered neurological function, but the right pelvic limb had developed a progressive paresis within the last 4 wk. Further diagnostics were not performed.

Case 3

A 7-year-old male intact pug dog was presented with a 10-month history of slowly progressive difficulty walking on both pelvic limbs. In addition, he was occasionally fecally incontinent. The neurological examination revealed an ambulatory paraparesis and mild pelvic limb ataxia. The response to the paw replacement test was reduced in both pelvic limbs whereas segmental spinal reflexes were normal in both pelvic limbs. There was no pain on spinal palpation. Signs were consistent with a T3–L3 spinal cord lesion. Lumbar myelography revealed a dilatation of the dorsal subarachnoid space from the middle of the T12 vertebra reaching caudally up to the caudal aspect of the L2 vertebra. The subarachnoid diverticulum resulted in a reduction of the normal spinal cord height by 50.0%. In addition, there was a mild IVD protrusion at the T12–13 disc space. Cerebrospinal fluid analysis revealed a nucleated cell count of 4 cells/μL (normal: ≤ 5 cells) and a total protein concentration of 0.84 g/L (normal < 0.40 g/L). Cells were differentiated as 55% lymphocytes, 36% monocytes, and 9% neutrophils.

Computed tomography confirmed a dorsal localization of the diverticulum. A sagittal T2-weighted MR image was obtained using the same settings as described. Intramedullary signal hyperintensity extended over the entire length of the T11 vertebra. Dilatation of the dorsal subarachnoid space was not definitely identified on the sagittal T2 weighted MRI.

A left-sided hemilaminectomy from T11 to T13 combined with a durectomy extending over the entire length of the hemilaminectomy opening was performed. A durectomy over the entire length of the diverticulum up to L2 was not attempted since dilatation of the subarachnoid space caudally was considered to be most likely secondary to any pathology at the cranial border of the diverticulum. The defect was covered with a thin layer of a moistened gelatin sponge. In addition, fenestration of the IVD T12–13 was carried out. Closure was done in a routine fashion. Histopathologic evaluation of the resected dura showed fibrous tissue without inflammation or neoplastic cells. These findings were consistent with a diagnosis of arachnoid diverticulum. The dog changed from a moderate ambulatory paraparesis to a severe ambulatory paraparesis the day following surgery. He regained the presurgical neurological condition within the following 3 wk. On a 5-month reevaluation the dog was improved compared to its presurgical status. He showed a mild pelvic limb ataxia only. The fecal incontinence had resolved about 3 mo after surgery.

Case 4

A 1.5-year-old male intact pug dog was presented with a 10-week history of scuffing of pelvic limbs and occasionally knuckling over, causing skin lesions dorsally on both pelvic limbs. The dog did not seem to be in pain since it was still jumping and walking stairs as before. Signs were non-progressive. The neurological examination revealed ambulatory paraparesis with normal response to the paw replacement test and normal segmental spinal reflexes in all limbs. There was no pain on spinal palpation. Signs were consistent with a T3–L3 spinal cord lesion. Lumbar myelography revealed a drop-shaped dilatation of the dorsal subarachnoid space over the IVD space T6–7 resulting in the diagnosis of an arachnoid diverticulum (Figure 3B). Spinal cord height at the maximal diverticulum extension was reduced by 41% compared to the spinal cord height 1 vertebra caudally. A mild hemivertebra at T5 was the only other abnormality within 2 vertebral bodies of the lesion. Computed tomography confirmed dorsal localization of the diverticulum. A sagittal T2-weighted MR image was obtained using the same settings as described. Intramedullary spinal cord signal hyperintensity was seen between vertebrae T6 and T7 cranial and caudal to the lesion, whereas there was a normal spinal cord signal at the center of the lesion (Figure 3A). Differentiation between intramedullary edema, dilated arachnoid space, and epidural fat was not possible on those sagittal views. The arachnoid diverticulum, however, could be identified retrospectively comparing myelography and MRI (Figures 3A, B).

Figure 3.

Sagittal T2-weighted spinal MRI (A) and lateral myelography (B) of the thoracic spine of a 1.5-year-old pug dog. There is an intraspinal signal hyperintensity on MRI (A). The arachnoid diverticulum at the cranial border of T7 (arrow) visible on the mylography cannot clearly be identified on the MRI.

A left-sided hemilaminectomy from T5 to T8 combined with durectomy was performed. Histopathologic findings on removed tissue showed the same changes as described in Case 3. The dog was neurologically improved 1 d after surgery and had a normal neurologic examination on re-checks at 3 wk and 4 mo.

Case 5

A 3.5-year-old female pug dog was presented with a 5-week history of scuffing both pelvic limbs, resulting in dorsal digital skin lesions. Signs had not improved following the administration of meloxicam (Metacam; Boehringer Ingelheim Vetmedica GmbH, Ingelheim, Germany), 0.1 mg/kg BW, PO, q24h for 1 wk. The neurological examination revealed ambulatory paraparesis and moderate pelvic limb ataxia. The response to the paw replacement test was absent in both pelvic limbs and segmental spinal reflexes were normal. There was no pain on spinal palpation. Signs were consistent with a T3–L3 spinal cord lesion. Lumbar myelography revealed a focal dilatation of the dorsal subarachnoid space at T8, reducing the spinal cord dorsoventral height at the maximum diverticulum extension by 37% compared to the normal spinal cord diameter 1 vertebra caudally (Figure 4). A hemivertebra at T5 was the only other abnormality in the immediate vicinity of the lesion. Computed tomography confirmed dorsal localization of the diverticulum. Surgery was declined by the owner. Medical treatment was initiated using a tapering anti-inflammatory dose of prednisolone (Prednisolon 5 mg; CP-Pharma, Burgdorf, Germany), starting with 0.36 mg/kg BW q12h. The dose was reduced by 50% every 2 wk and prednisolone was discontinued after 6 wk. There was no change in neurologic status for the next 10 mo.

Figure 4.

Lateral myelography of the thoracic spine of a 3.5-year-old pug dog with a dorsal arachnoid diverticulum at T8, which does not have a drop-shaped appearance.

Discussion

The rate of diagnosed arachnoid diverticula in pug dogs during the 4-year period investigated here was 38% of all pug dogs with T3–L3 spinal cord lesions, whereas only 13% of pug dogs were diagnosed with thoracolumbar IVD herniation during the same time. Therefore, arachnoid diverticula should be considered one of the major differential diagnoses in pug dogs presented for a chronic, non-painful T3–L3 spinal cord lesion. However, this finding does not correspond with previously reported data. There were only 3 pug dogs among the previously reported 88 canine cases of arachnoid diverticula resulting in a rate of 3% (1–19). In contrast, in our hospital, 83% of the entire population of dogs with arachnoid diverticula were pug dogs. There was only 1 dog of another breed, a Rhodesian ridgeback, with an arachnoid diverticulum during the observation period of 4 y. The cause of the increased rate of diverticula in pug dogs in our study compared to the literature is unknown. One explanation may be the increasing popularity of pug dogs. The annual number of pug dog puppies registered with the German Kennel Club increased by 317% between 1998 and 2010, whereas the annual number of puppies of all breeds decreased by 20% during the same time (21).

The majority of canine subarachnoid cysts reported in dogs were located in the cervical area (63%) (1–19). In small breed dogs, however, subarachnoid cysts were nearly exclusively located in the thoracolumbar area (91%) (1–19). Small breed dogs are usually affected at an older age (17). Similarly, none of the 5 pug dogs presented here was < 1.5 y, with a median of about 4.5 y, suggesting acquired rather than congenital arachnoid diverticula. Alternatively, a congenital diverticulum may cause chronic spinal cord compression resulting in neurological signs later in life once the spinal cord compensatory capabilities and spinal cord plasticity have been exhausted. This process may even be accelerated if an initially smaller cyst increases in size over time. None of the dogs in this study experienced obvious spinal pain. Another study found spinal pain in 4/14 dogs with spinal arachnoid diverticula, whereas spinal pain was detected in only 1/10 dogs with thoracolumbar spinal arachnoid cysts in another study (15,17). In contrast, neuropathic pain is a common feature affecting 45% of human patients with intradural arachnoid cysts, the same type of cysts seen in dogs (22). This pain is usually experienced by humans as radicular pain indicating some nerve root involvement (22). The underlying pathophysiology explaining the difference in the frequency of spinal pain between humans and dogs is unknown. Urinary and/or fecal incontinence is another common clinical feature in affected dogs and humans (15,17,22). It seems to be more common in dogs with thoracolumbar than in those with cervical cysts, but only one of the dogs reported here had fecal incontinence (17).

All diverticula found in the 5 dogs reported here were located between T6 and L2, with only 2 of them showing the typical drop-shaped appearance and the other 3 dogs having a focal arachnoid dilatation (Figures 1,4). Additional spinal pathology such as hemivertebra, mild disc protrusions, or spondylosis deformans in the immediate vicinity of the arachnoid diverticulum was seen in 4/5 dogs. It may raise the suspicion that those diverticula developed secondary to other primary spinal lesions as has been suggested previously. Acquired subarachnoid cysts have been reported in dogs in association with arachnoiditis (1,16), hemivertebra (2), intervertebral disc disease (12,15,17), scoliosis (9), and spinal trauma (17). Arachnoiditis as underlying etiology seems to be unlikely in the dogs presented here, at least in the 2 dogs in which histopathology failed to detect inflammatory changes in the removed dura mater. It is uncertain if the existence of hemivertebrae or mild disc protrusions were in causative relationship to the spinal diverticula in the dogs reported here since many pug dogs have similar lesions without developing arachnoid diverticula. However, careful screening of spinal images for any potential underlying pathology should be done since surgically addressing the diverticulum alone may not be sufficient in some dogs. Additional decompression of mild disc protrusions or stabilization in cases of hemivertebrae associated minor instability may be indicated in some cases.

Surgical options for removal of arachnoid diverticula include durotomy/durectomy and marsupialization, with the majority of dogs reported so far having received the latter (5,11,15,16). Comparison of outcomes from both surgical techniques in dogs is limited to 1 study of 9 dogs with marsupialization and 6 dogs with durotomy (17). The authors concluded that there might be a slight tendency towards better outcome after marsupialization. Surgery results in a good short-term outcome in most cases, even though many dogs do not become neurologically normal after surgery (15,17,18). Short-term prognosis seems to be less favorable in dogs which have multiple cysts (15). Long-term outcome, however, is affected by a recurrence rate of clinical signs in up to 30% of cases (17).

Prognosis in the 5 pug dogs reported here was not as favorable as reported in the literature with 3/4 dogs experiencing neurological deterioration immediately after surgery. It took those 3 dogs up to 8 mo to recover presurgical neurological function and to improve beyond the presurgical status. Only 1 dog showed immediate postsurgical improvement and complete resolution of neurological deficits within 3 wk. Hence, prognosis may be different in pugs than in other breeds. Interestingly, 1 pug dog that was treated medically using an anti-inflammatory dose of prednisolone was neurologically stable for a follow-up time of 10 mo.

The postsurgical deterioration and lack of complete resolution of clinical signs in the long-term may raise the question if clinical symptoms may have been caused by a degenerative spinal cord disease or by a combination of such a disease and the arachnoid dilatation rather than by the dilatation alone. Degenerative myelopathy has been seen in pug dogs (personal communication, Dr. Joan R. Coates, University of Missouri, 2012) and should therefore be considered an additional differential diagnosis in pug dogs with chronic-progressive paraparesis. However, the final improvement of the neurological deficits beyond the presurgical condition in the pug dogs reported here, though requiring several months, does support the assumption that the arachnoid diverticula were mainly responsible for the neurological deficits observed in those dogs. Postsurgical deterioration is most likely caused by the chronic course of the disease resulting in significant axonal loss, making the spinal cord vulnerable to any manipulation during surgery. In addition, reperfusion injury following spinal cord decompression may have added to the deterioration immediately after surgery.

Arachnoid diverticula should be considered as a major differential diagnosis in pug dogs with progressive T3–L3 spinal cord lesions. Spinal pain and urinary or fecal incontinence are not typical features of thoracolumbar diverticula in pug dogs. Diverticula can be visualized on myelography or CT-myelography. They may be focal dilatations rather than the typical drop-shaped structures seen in other breeds. Images should be carefully screened for associated spinal lesions that may have to be addressed as well. Short-term prognosis after surgical therapy may not be as good as reported for other canine breeds. For the long-term, however, dogs improve beyond the presurgical status, though they may not become neurologically normal. CVJ