Syringohydromyelia in horses: 3 cases

Abstract

Syringomyelia and hydromyelia are cavitary lesions of the spinal cord that may be acquired or congenital. These lesions are not frequently reported in large animal species. The presenting complaints, clinical, gross pathological, and histopathologic findings of 2 cases of syringomyelia and 1 case of hydromyelia in horses are described.

Syringomyelia and hydromyelia are abnormal fluid-filled cavitary lesions of the spinal cord (1,2). They have been reported in domestic species, such as horses (3,4), cats and dogs (5), and exotic species, including a Reeves’ Muntjac (6), and a neonatal dromedary (7). With advances in imaging techniques, a clinical diagnosis of syringomyelia is being made more commonly in small animals (1); however, there are few clinical reports in horses (3,4). Here we report the clinical, gross necropsy, and histopathologic findings of 2 young horses diagnosed postmortem with presumed congenital syringohydromyelia and 1 adult diagnosed with putative acquired syringomyelia.

Syringomyelia is characterized by a cavitary space (syrinx) within the spinal cord that extends through one or several vertebral segments (1). The syrinx contains fluid that is consistent with the characteristics of extracellular and cerebrospinal fluids (2). Depending upon location of the cavitary space within the spinal cord, the term “hydromyelia” is used for dilatation of the central canal, as determined by histologic examination and presence of ependymal cells lining the canal, and “syringomyelia,” for fluid-filled cavities within neuroparenchyma usually lined by glial cells (1). Syringohydromyelia is a global term that has been proposed since hydromyelic cavities often include a contiguous pocket of neuroparenchyma lined by glial cells and syrinxes within neuroparenchyma (syringomyelia) often rupture into the central canal (1,2,5). In horses, syringohydromyelia and syringomyelia have been histologically confirmed previously; however, to the authors’ knowledge, hydromyelia has not previously been described in the horse (3,4). Clinical signs associated with syringohydromyelia reflect the affected neuroanatomic area of the lesion (5). In dogs, syringomyelia most commonly involves the water-shed zones more centrally and dorsally located within the spinal parencyma in the dorsal funiculi (8). The presence and/or expansion of the cavity often results in progressive deficits that may include pain, loss of sensation, paresis, and autonomic dysfunction (9).

Case descriptions

Case 1

A 6-day-old, male, Appaloosa foal was evaluated by the Iowa State University Veterinary Medical Center (ISU VMC) emergency service because of diarrhea and progressive lethargy of 1-day duration. Following an uneventful parturition, a slight right scoliosis and intermittent spastic extension of the right thoracic limb were noticed by the owner; the scoliosis had slowly resolved by 4 d of age. The foal received 1 dose of enrofloxacin and dipyrone for the diarrhea before referral.

On presentation, the foal was depressed, weak but ambulatory, febrile (102.5°F; reference interval: 99.0 to 101.8°F), tachypneic [70 beats/min (bpm); RI: 30 to 40 bpm], and was approximately 5% dehydrated. Mild tendon laxity was evident in all limbs. The foal passed moderate amounts of watery yellow feces and ground its teeth occasionally. A small amount of urine dripped from the umbilical stump when the foal urinated. All other physical examination findings were normal. Hematology, serum biochemistry, and blood gas analysis revealed a toxic left shift, mild electrolyte derangements, mild hypoalbuminemia, and a mild metabolic acidosis. The foal received 2 L of plasma, intravenous (IV) crystalloid fluids, antimicrobials, non-steroidal anti-inflammatory medication, and supportive care.

The diarrhea and attending clinicopathologic abnormalities gradually resolved over the next 4 d. During hospitalization, the foal intermittently exhibited spastic extension of the right thoracic limb, especially when excited or rising from recumbency. The neurologic signs gradually worsened to include intermittent bilateral thoracic limb tremors. At a trot, the foal also exhibited occasional spasticity in the hind limbs, manifested as “bunny hopping.” Cervical radiographs (day 2) showed no abnormalities. Based on the neurologic examination, a neuroanatomic localization to the caudal cervical-T3 spinal segments was derived with a congenital myelodysplasia or other congenital defect as a likely cause. The foal was discharged after 4 d of hospitalization with instructions to continue antimicrobials for an additional 8 d for treatment of the patent urachus. Six months later, the foal was presented to the ISU VMC for euthanasia and necropsy examination due to progressive worsening of previously identified neurologic signs. At this time it was noted that the foal had reduced growth in the thoracic girdle and thoracic limbs compared with the pelvic region and limbs. A complete necropsy was performed and findings relevant to the neurologic deficits included a large cavitary lesion 2.5 cm in length along the C7-T1 segments of the gray matter (syringomyelia). The syrinx involved the cervical cord lateral to the central canal and resulted in obliteration of the central canal at the distal most portion of T1 and dorsoventral flattening of the C7 segment (Figure 1). Moderate stenosis of the spinal canal at C7 was observed grossly. Histopathologically, the spinal nerve roots at C7 were associated with sheets and clusters of glial cells. No gross or histopathologic lesions were noted in the remainder of the brain and spinal cord.

Figure 1.

A — C7 segment, cervicothoracic spinal cord, 6-month-old Appaloosa colt (case 1). The grey and white matter of the spinal cord (sc) are severely compressed and thinned due to syringomyelia. The arachnoid layer (a) of the meninges is expanded by fluid. Dura mater (d). Hematoxylin and eosin. Bar = 1 mm. B — Slide of the 6-month-old Appaloosa colt (case 1). Showing that the architecture of the spinal cord is distorted due to release of excessive fluid.

Case 2

A 3-month-old Paint filly was presented for bilateral thoracic limb lameness. Physical examination findings were within normal limits aside from significant bilateral wear of the toe region of the thoracic limbs. Neurologic examination revealed normal mentation and cranial nerves. When lead, the filly had a mildly spastic gait. She appeared to have diminished or weakened extensor function bilaterally in the thoracic limbs manifest by frequently dragging her toes and flipping her feet out in front of her during the end of the swing phase of protraction. The filly stumbled when the cranial phase of protraction was decreased. The filly could negotiate a curb; however, to navigate the transition to an elevated surface, she had to shift her weight to the pelvic limbs and exuberantly extend the thoracic limbs. Proprioception of the thoracic limbs was difficult to assess at the walk as a mild degree of thoracic limb paresis affecting the cranial phase of the stride was also suspected. Radiographs of the left shoulder were within normal limits, making a bilateral shoulder problem unlikely. Proprioception of the pelvic limbs was normal. Based on these findings, a neuroanatomic localization to the C1-T3 spinal cord segments was made. A congenital myelodysplasia and equine degenerative myeloencephalopathy with sparing of peripheral tracts were considered as possible etiologies. The owners elected to wait 8 mo to see if the deficits improved with time. Reevaluation showed a static neurologic condition with otherwise normal growth. The owners elected for euthanasia and complete necropsy.

Necropsy findings pertinent to the neurologic deficits included hydromyelia (C1-C6) with maximal dilation of 1.3 mm at segment C4-C5, while proximal and distal segments of the cervical cord were mildly dilated at 0.75 to 0.80 mm. No obstruction to CSF flow or cause for increased CSF production, such as a space occupying lesion, was observed at necropsy or by histologic examination (Figure 2). There was, however, a mild stenosis of the vertebral canal at C6-C7. The cerebrum, cerebellum, and ventricles were of normal size and morphology. The brachial plexi were symmetric and histologically normal.

Figure 2.

A — Photomicrograph, C4-C5 segment, cervical spinal cord, of the 11-month-old filly (case 2). The central canal is moderately dilated. Blood vessels in the surrounding gray horns (gh) and dorsal funiculi (df) are congested. Hematoxylin and eosin. Bar = 1 mm. B — Photograph of microscope slide showing C4-C5 segment, cervical spinal cord, 11-month-old filly (case 2). Bar = 1 mm.

Case 3

A 4-year old grade pony in good flesh was presented with a history of acute onset of intermittent kicking of the dorsal aspect of the right metatarsal III by the left pelvic limb. This self-mutilating intermittent behavior had begun 2 d prior to presentation; however, the frequency of this action had increased over this time frame resulting in an extensive, severe, open wound to the right MtIII with exposure of bone. No known traumatic insult, such as a fall, was observed; however, the pony had been pastured with other new horses for 3 d prior to the onset of signs. The pony appeared apprehensive moments before an episode of self-mutilation appeared, suggesting awareness of an unusual sensation. The activity of the left pelvic limb appeared to be involuntary with damage to the right hind occurring while the pony tried to maintain its posture. The owner was concerned about the pony’s immediate quality of life and elected for euthanasia and necropsy without further treatment or diagnostics. Neuroanatomic localization suggested prosencephalon (seizure activity) and/or a spinal cord lesion involving the pelvic spinal segments, left > right.

Necropsy findings relevant to the pony’s behavior included a 24 × 3 cm laceration of the skin and underlying tissue on the dorsomedial aspect of the right metatarsal region. A 2-cm indentation was noted on MtIII. At C2-C3, a focal area of margination of lymphocytes and plasma cells was present within blood vessels of the gray matter with rare foci of lymphoplasmacytic cuffing. At T13-L1, a 2 × 2 mm cavitary lesion was present in the ventral funiculus of these segments (Figure 3). The brain, remaining spinal cord segments, skeletal muscle of the left and right caudal thighs, and sciatic nerves were within normal limits. None of the lesions detected at necropsy could be easily interpreted as causative for the abnormal behavior from a neuroanatomic viewpoint. However, given that no clear causal association was evident, we speculate that the T13-L1 syrinx may have impacted segments further caudally via hydrodynamic influences possibly causing pain, dysesthesia, or paresthesia of the distal limb.

Figure 3.

A — Photomicrograph, T13-L1 segment, spinal cord, 4-year-old grade pony gelding (case 3). Bar = 200 μm. B — Photograph of microscope slide showing T13-L1 segment, spinal cord, 4-year-old grade pony gelding (case 3). Bar = 1 mm.

Discussion

Clinical signs of syringohydromyelia reflect the location of the lesion in the spinal cord. Although no pathognomonic signs are associated with syringohydromyelia, paresthesia, and dysesthesia in conjunction with signs of spinal cord disease should prompt consideration of syringohydromyelia. In addition, acute exacerbation of signs of spinal cord disease during changes in posture or activity is suggestive of syringohydromyelia. An association of scoliosis with syringohydromyelia has previously been identified in other species and should increase the index of suspicion of syringohydromyelia in horses as well. A history of trauma in conjunction with signs of spinal cord disease is suggestive of acquired syringohydromyelia, whereas signs of spinal cord disease observed shortly after birth may indicate congenital syringohydromyelia. In this report, the lesions observed at necropsy for cases 1 and 2 were believed to have been congenital, as the neurologic deficits were present at a young age and there was no known trauma at birth or during the neonatal period. In contrast, case 3 was 4 years old at the onset of clinical signs, supporting acquired development of the syrinx, possibly associated with a traumatic episode.

Dilatation of the central canal can communicate directly with the 4th ventricle or it can arise below a syrinx-free segment of the spinal cord (9,10). In humans, the former is more common (9). In dogs, it has been reported that dilatation of the central canal can directly communicate with the 4th ventricle without concomitant hydrocephalus (5,11). Interestingly, in 1 of the horses in this report (case 2), the syrinx appeared to communicate with the 4th ventricle and there was no hydrocephalus.

Based upon clinical and diagnostic imaging techniques, it is difficult to differentiate between syringomyelia and hydromyelia (5,6). Histology is generally required for a definitive diagnosis. Hydromyelia, as reported here, may be further classified by whether or not there is association with the 4th ventricle (5). A syrinx of the central canal that communicates with the 4th ventricle is found with hydrocephalus that results from obstruction of CSF flow distal to the outlets of the 4th ventricle (6). In contrast, hydromyelia occurring distal to a syrinx-free segment of the spinal cord is considered to be non-communicating with the 4th ventricle. In the latter instance, pathology of the spinal subarachnoid space hampers normal CSF flow, resulting in fluid being forced into the central canal through the interstitial spaces of the spinal cord (6). In humans, noncommunicating central canal syrinxes are associated with Chiari I malformation, spinal arachnoiditis, basilar impression, occipital encephalocele, congenital syrinx of the terminal ventricle, and cervical spinal stenosis (10). In case 2, the spinal canal at C6-C7 was mildly narrowed and the central canal was of normal diameter in spinal segments beyond this point. It is not clear that mild cervical vertebral body stenosis was the cause of syrinx formation. Indeed, there was no structural defect within the parenchyma of the cord that would have resulted in obstruction of flow of CSF. Despite the observation that mild cervical vertebral body stenosis co-localized with the extent of dilatation of the cervical central canal, the central canal was dilated throughout the cervical region, including spinal segments C1-C2. Absent serial cross-sectional measurements of the central canal cranial to C1, it appears that the dilatation of the central canal communicated directly with the 4th ventricle. As a result, case 2 was classified as having a communicating central canal syrinx. In humans, a characteristic feature of communicating syringomyelia is its association with hydrocephalus (10). In case 2, however, this relationship was not observed as there was no evidence of hydrocephalus. Similarly, in dogs, hydromyelia associated with hydrocephalus appears to be uncommon, though it has been reported (11). Paracentral parenchymal changes were observed throughout the C1-C6 cervical cord region in case 2 and were characterized by multifocal hemorrhages, distortion of the grey matter, rare swollen axonal sheaths containing rare gitter cells, and congestion of blood vessels.

In humans, communicating syringohydromyelia has been associated with non-specific segmental neurological deficits, including spastic weakness of the lower extremities, hyperreflexia, impaired balance, weakness of the upper extremities, and numbness of the lower extremities (10,12). Case 2 appeared to have weakness of the extensors of the thoracic limbs as evidenced by difficulty advancing the limbs, bilateral wear of the toes of the forelimbs, and difficulty walking over raised objects (curb). However, proprioception did not appear compromised as placement when stepping down was normal, and postural and placement responses were normal. Nonetheless, the abnormal cranial “swing phase” of the stride made assessment of proprioception difficult.

Etiology for syrinx formation within either the central canal or spinal cord neuroparenchyma generally falls under acquired or developmental causes. Hydromyelia involving few spinal cord segments tends to be found in older children and adults and is often associated with trauma, whereas larger defects, frequently extending throughout the length of the cord, are associated with congenital lesions or early childhood disease (10). In cases of injury or trauma, syrinxes tend to be located in the central gray matter or dorsal and lateral white matter, possibly as a result of changes in vascular distribution (“watershed zones”) (13). In case 1, 3 segments (C7-T1) were affected. In case 2, the hydromyelia was maximal at segments C3-C4 and C4-C5 while proximal and distal segments of the cervical cord were mildly dilated, consistent with reports in humans where the hydromyelia was suspected as being congenital.

Several pathophysiologic mechanisms have been proposed to explain development of hydromyelia and syringomyelia; however, none adequately explains all cases (5,10,14). In most cases, however, pathologic causes for alterations in the flow of CSF can be detected. It is believed that this disturbance in CSF flow leads to increased ventricular pressure and consequent hydrodynamic increases in CSF pressure within the central canal (hydromyelia) or cavitation of the spinal cord parenchyma (syringomyelia). Causes of syringohydromyelia can generally be categorized by obstruction of flow of CSF, constrictive lesions of the spinal cord, or space occupying lesions involving the spinal cord (2). In communicating syringohydromyelia, caudal fossa masses are relatively common, and frequently result in hydrocephalus (2). Suspected causes of syrinx formation in dogs and cats include obstructive lesions to the lateral apertures, abnormal changes in hydrodynamics between the brain and spinal cord at the foramen magnum, developmental defects of the spinal cord, inflammatory insults to the cord, and stenosis of the central canal (5,8). Defects of the foramen magnum have been previously reported in animals (8). In large animals, the occurrence is most frequently detected in calves. The most common malformation noted is caudal extension of the cerebellar vermis into the cranial aspect of the vertebral canal (8). Dandy-Walker-Like Syndrome has previously been reported in a neonatal foal (15). In that report, a meningocele was associated with dilatation of the 3rd and lateral ventricles, as well as the mesencephalic aqueduct; however, the dilatation of the mesencephalic aqueduct did not extend caudally to include the central canal (15). In many cases, obvious pathologic causes are not found, leading to hypotheses based on hydrodynamic theory to explain the development of both syringomyelia and hydromyelia (14). These approaches view development of syringohydromyelia as a consequence of changes in pressure gradient between CSF in the subarachnoid space and pulse pressure within the compliant spinal cord. The pressure gradient is postulated to be generated by the Venturi effect, as explained by the Bernoulli theorem: the total mechanical energy of flowing fluids remains constant, meaning that the increased fluid velocity at a narrowed flow channel decreases the pressure in the fluid. Thus, an increase in CSF velocity at the site of an obstruction (subarachnoiditis, vertebral stenosis) paradoxically decreases the pressure in the CSF compared to that in the adjacent spinal cord (14,16). As a result, the cord is distended by ECF at and distal to the subarachnoid impediment (14).

The increase in pressure associated with development of syringohydromyelia is likely dynamic. Brain volume increases in response to cardiac systole and decreases during diastole (5). As a result of intracranial and CSF fluxes in volume, during cardiac systole, CSF flow normally increases through the foramen magnum to the cervical spinal cord and reverses during cardiac diastole. Lesions which interfere with flow may result in CSF accumulation in the ventricles of the brain (hydrocephalus) or within the cord (syringohydromyelia). Transient increases in intracranial, intrathoracic, or intraabdominal pressure may contribute to development of syringohydromyelia. While no studies have been performed in clinical veterinary patients which demonstrate any contribution from changes in CSF pressure gradients to the development of syringohydromyelia, exacerbation of clinical signs in dogs has been reported during exertion, barking, or excitement (17). Such changes in CSF pressure gradients would be expected to occur with changes in body posture, abdominal press against a closed glottis (grunting), coughing, sneezing, etc., and to alter subarachnoid pressure. In case 1, the unusual onset of right forelimb extension associated with rising from recumbency may have been related to increased subarachnoid pressure associated with the change in posture, resulting in acute exacerbation of increased subarachnoid pressure-related clinical signs. Similarly, the occasional pelvic limb spasticity detected during excitement may have been related to CSF hydrodynamics and increased pressure. In contrast, none of the lesions detected at necropsy of case 3 could be easily interpreted as causative for the abnormal behavior. However, given that no clear causal association was evident, we speculate that the T13-L1 syrinx may have impacted segments further caudally via hydrodynamic influences. Experimental studies that support this notion include sheep and rat models in which canalicular syringomyelia occurs secondary to arterial pulsation-driven perivascular flow of CSF into spinal cord parenchyma with resultant extracanalicular formation of a syrinx. Suspected inciting causes include trauma, hemorrhage, and excitotoxic events (18) with active perivascular flow contributing to enlargement of the syrinx. Post-traumatic syringomyelia is speculated to be the result of initial syrinx formation, with arachnoiditis causing an increase in subarachnoid space pulse pressure and reduced compliance (18).

Clinical signs associated with syringohydromyelia reflect the affected neuroanatomic area of the lesion. With lesions involving cervical segments, ataxia and paresis are frequently detected and are usually more severe in the pelvic limbs. This condition is referred to as central cervical spinal cord syndrome, generally occurs following cervical trauma, and is related to the somatotopic organization of tracts coursing in the spinal cord (17,19,20). The observation of neurologic deficits which are worse in the thoracic limbs may be explained by the fact that the spinal tracts which project to the thoracic limbs are more centrally located in the cord and therefore more severely affected by the syrinx than the more peripherally located tracts which project to the pelvic limbs (17,19,20). The bilaterally symmetric neurologic deficits in case 2 were confined to the thoracic limbs and histopathology confirmed a C1-C6 syrinx of the central canal. Together with the mild stenosis of the vertebral canal at C6-C7, we considered central cervical spinal cord syndrome as possible, although no known traumatic incident to the cervical region was known to have occurred.

Scoliosis is a frequent clinical finding among dogs and humans with syringomyelia (5,17). It is believed that expansion of a syrinx damages ventral gray horn cells which innervate the limb and paraspinal musculature. Lower motor neuron signs that may consequently manifest include muscle atrophy, paresis, and decreased spinal reflexes. If damage to the ventral gray columns is asymmetric and severe, an ipsilateral convex scoliosis may arise (5,17). In case 1, a cervical scoliosis was apparent after birth, but improved over a few days. Together with the persistent, intermittent clinical finding of right thoracic limb spasticity, the scoliosis supported an asymmetric cervical to T3 lesion.

This report summarizes findings of presumed congenital syringomyelia and hydromyelia and 1 case of suspected acquired syringomyelia in the horse. With more widespread availability of magnetic resonance imaging and contrast computed tomography, we speculate that these lesions will be increasingly recognized in the horse.