C.W. Dewey, DVM, MS, DACVIM (Neurology), DACVS
Caudal occipital malformation syndrome (COMS) is the canine analog of Chiari type I malformation of people. Although only recently described in dogs, COMS is a very common neurologic disorder in this species. This disease is almost exclusive to small breed dogs, with the Cavalier King Charles spaniel (CKCS) being the most over-represented. There is convincing evidence in the CKCS breed that COMS is a heritable disease, although the exact mode of inheritance has not been determined. The disorder is a congenital malformation of the caudal occipital region of the skull, leading to overcrowding of the caudal fossa and compression of the cervicomedullary junction at the level of the foramen magnum. Most dogs with COMS have syringomyelia, an accumulation of fluid within the spinal cord, as a consequence of the malformation. In patients with COMS, there tends to be some level of cerebellar compression as well as constriction of the cervicomedullary junction in the vicinity of the foramen magnum. With chronic bony compression at the cervicomedullary junction and probable turbulent CSF flow and pressure changes in this region, it is thought that the underlying meninges become hypertrophied with time. In both humans with Chiari type I and dogs with COMS, there is pathological evidence of dural fibrosis in the region of the malformation. In COMS, as in Chiari type I of people, the caudal aspect of the cerebellum is often projecting into or through (herniation) the foramen magnum, contributing to obstruction of CSF flow between intracranial and spinal compartments.
Progressive alterations in pressure dynamics between the intracranial and spinal compartments are believed to be responsible for the development of clinical signs of COMS. Although aberrant pressure dynamics due to obstruction of CSF pathways at the level of the foramen magnum are generally agreed to cause syringomyelia in COMS, the exact mechanism of this development is unknown and there are multiple theories proposed to explain it. Many of these theories operate on the probably incorrect premise that the syringomyelia fluid is CSF which is forced into the central canal of the spinal cord. The newer theories suggest that the syringomyelia fluid is actually derived from extracellular fluid from the cord itself, either driven into the central canal via a pulse pressure wave from behind the foramen magnum obstruction and/or drawn into the central canal via a centrifugally directed hydrostatic pressure force within the spinal cord. In normal dogs, there is pulsatile CSF flow across the foramen magnum from intracranial subarachnoid space to cervical spinal subarachnoid space and back again during systole and diastole, respectively. With an obstruction at the foramen magnum as occurs with COMS, CSF does not flow well in either direction. In this scenario, the pressure exerted during systole may drive either CSF or a pressure wave from the intracranial compartment into the central canal region of the cranial cervical spinal cord, causing it to progressively expand. This has been referred to as the "water-hammer" effect. Another theory proposed is that CSF is "sucked" into the central canal region of the cervical spinal cord, especially during maneuvers that lead to sudden increases in intrathoracic an intraabdominal pressure (e.g., coughing, sneezing, exercising). These Valsalva maneuvers lead secondarily to increased intracranial and intraspinal pressure via epidural venous distension. Because intracranial pressure is higher than in the cervical cord region, CSF fluid is drawn into the cervical cord when there are rapid increases in pressure. Pressure within the spinal compartment tends to increase more rapidly in the lumbar versus cervical regions, further promoting CSF movement into the cervical cord via this "suck" effect. The "slosh" phenomenon may also be involved in expansion of a syrinx. With distension of epidural veins during Valsalva events, CSF flows more freely within the syrinx than in the compressed subarachnoid space. Therefore, sudden CSF pressure waves cause CSF within the syrinx cavity to "slosh" around, fissuring surrounding parenchyma and enlarging the syrinx. The combination of spinal epidural vein distension (and resultant pressurization of the subarachnoid space) and obstruction to CSF flow from the cervical spine to the intracranial compartment may also result in forcing subarachnoid CSF down perivascular spaces into the spinal cord parenchyma, progressively enlarging the syrinx. It has also been proposed that the displaced caudal cerebellum acts like a "piston". This theory suggests that the displaced cerebellum moves further caudally during systole, obstructing the subarachnoid space at the foramen magnum and exaggerating the systolic pulse pressure wave that is transmitted from intracranial to spinal compartments; this further forces CSF through perivascular spaces into the syrinx. Although all of the above theories may contribute to the development of a syrinx, none of them are adequate explanations for this phenomenon. It has been shown that the syrinx usually has a higher pressure than the subarachnoid space, which would argue against theories that propose CSF is being forced or sucked into a low pressure system form a higher pressure system. In addition, syrinx fluid is not identical to CSF fluid; it has a lower protein concentration and is more consistent with extracellular fluid. Several related theories have been proposed that are more likely to adequately explain the pathogenesis of syringomyelia formation with COMS or Chiari type 1 malformation. The "intramedullary pulse pressure" theory proposes that the spinal cord parenchyma distal to the foramen magnum compression is subjected to distending forces that tend to pull the tissue in an outward or centrifugal direction. The combination of transmittal of the systolic pulse pressure wave to the spinal cord parenchyma (due to obstruction of the subarachnoid space) and decreased subarachnoid space pressure in the spinal cord region (due to obstruction of the subarachnoid space rostral to the foramen magnum) lead to this mechanical distension. Over time, the distension leads to a cavity formation (syrinx), which is filled with extracellular fluid. The "Venturi effect" describes a similar mechanical spinal cord distension caused by increased CSF velocity distal to an obstruction. The obstruction (i.e., foramen magnum occlusion in COMS) causes a narrowing of the subarachnoid space and a resultant increased fluid velocity distal to the obstruction. This increased velocity lowers the hydrostatic pressure, producing a centrifugally directed suction effect, leading to spinal cord distension. This theory also assumes that the accumulated fluid in the syrinx is extracellular fluid and at a higher pressure than the subarachnoid space. Finally, there is a "vascular" theory to explain the development of syringomyelia in COMS cases. With increased CSF pressure in the intracranial compartment vs. the spinal compartment due to foramen magnum obstruction (especially during Valsalva maneuvers and systole), the venous and capillary beds become collapsed in the intracranial region and distended in the cervical spinal cord region (caudal to the obstruction). This occurs because CSF and venous pressure normally remain closely matched in both cranial and spinal compartments, and the venous system does not become obstructed as does the subarachnoid space with foramen magnum obstruction. With foramen magnum compression the transmural pressure (difference between intravascular and interstitial pressure) of the venous and capillary system on either side of the obstruction is no longer uniform throughout the spinal cord. The uneven vascular expansion and contraction that ensues causes damage to the surrounding spinal cord. This hydrostatic stress-mediated damage to the spinal cord disrupts the blood-spinal cord barrier, promoting the accumulation of extracellular fluid within the spinal cord (i.e., syrinx development). Common to all theories of syrinx development in COMS is the causative factor of obstruction of normal CSF flow at the foramen magnum.
Clinical Features of COMS in the CKCS Breed
Most dogs with COMS are presented for evaluation as young adults, between 3 and 6 yrs of age. However, the age range for this disorder is very broad, ranging from less than 6 mos to more than 12 yrs of age. Dogs that are presented at less than 2 yrs of age often have more severe clinical signs than older dogs. Similar to Chiari type I of humans, there is a wide spectrum of possible neurologic presentations for dogs with COMS, including cervical myelopathy, cerebellovestibular dysfunction, and forebrain dysfunction (e.g., seizure activity). By far, evidence of cervical dysfunction and cerebellovestibular dysfunction are the most common and are often both present (e.g., multifocal CNS disease). Most of the COMS cases that the author encounters are presented for signs referable to the cervical region (e.g., neck pain, scratching activity) and subtle signs of central vestibular dysfunction are apparent on neurologic examination. Occasionally, dogs with COMS and cervical syringomyelia present with a specific variant of cervical myelopathy called central cord syndrome. In this syndrome, the outwardly expanding syrinx causes more LMN damage to the thoracic limb musculature than white matter damage (to pelvic limbs); the result is thoracic limb paresis (often LMN in nature) that is notably worse than pelvic limb paresis. In some cases, the pelvic limbs may appear normal. Some specific clinical findings in dogs with COMS include cervical and cranial hyperesthesia, decreased menace responses with normal vision, positional ventrolateral strabismus, thoracic limb weakness, pelvic limb ataxia, persistent scratching (at the head, neck, and shoulder region-often without making skin contact), scoliosis, facial nerve paresis/paralysis (unilateral or bilateral), and hearing abnormalities. The persistent scratching activity and scoliosis are fairly unique clinical signs associated with syringomyelia. In the author's experience, these are more commonly encountered in the CKCS breed than in other breeds with COMS and syringomyelia. The scratching activity is believed to be due to the syrinx interfering with spinothalamic tracts and/or dorsal horn neurons, resulting in abnormal sensations (dysesthesia/paresthesia). Scoliosis (torticollis) is most likely due to asymmetric syrinx damage to sensory proprioceptive neurons innervating cervical musculature; an alternative, less likely hypothesis is syrinx damage to LMNs innervating cervical musculature. Scratching activity and neck discomfort often are exacerbated by abrupt weather changes, stress or excitement, and physical contact with the neck/shoulder region (e.g., collar). The presence of both pain and scoliosis has been shown to be significantly correlated with syrinx width in CKCS dogs with syringomyelia secondary to COMS. It is important to realize that, especially in the CKCS breed, other conditions may account for some of the clinical signs. An enigmatic ear problem of the CKCS breed, called primary secretory otitis media (PSOM) has been described. Clinical signs of PSOM include apparent pain around the head and neck area, scratching of the head and neck, facial paralysis, and head tilt. Idiopathic epilepsy is also a prevalent disorder in the CKCS breed. Seizures have been reported to occur in 10% to 12% of humans with Chiari type I malformation; in the author's experience, seizure activity is an infrequent concurrent occurrence in COMS cases, and it is usually not possible to distinguish whether this is due to COMS or concurrent idiopathic epilepsy. Congenital deafness is also well-described in the CKCS breed. The severity and rate of progression of COMS in dogs is variable, ranging from asymptomatic (i.e., finding evidence of COMS while imaging for some other reason) to extreme pain and debilitation with rapid worsening. In addition, some dogs with COMS have other concurrent disorders (e.g., disk extrusion, inflammatory brain disease) that could explain observed clinical signs. In such situations, it may be difficult to discern if the COMS is the main problem, contributory, or an incidental finding.
Diagnosis of COMS
Diagnosis of COMS is made by MR imaging. Magnetic resonance imaging is also the preferred imaging modality for diagnosing syringomyelia. The malformation is best visualized on a midsagittal view (preferably T2-weighted), which includes the caudal fossa and cranial cervical cord. Consistent findings on MR imaging indicative of COMS are attenuation/obliteration of the dorsal subarachnoid space at the cervicomedullary junction and rostral displacement of the caudal cerebellum by the occipital bone. Other common MRI findings in COMS include syringomyelia (usually C2 level caudally), herniation of the caudal cerebellum through the foramen magnum, and a "kinked" appearance of the caudal medulla. Phase-contrast MRI (cine-MRI) is often used to measure CSF flow in humans with Chiari type I malformation, and has recently been evaluated for use in dogs with COMS. Occasionally, dogs with MRI findings consistent with COMS will have evidence of other congenital disorders such as intracranial arachnoid (quadrigeminal) cyst, malformation of the C1 and or C2 vertebrae, and hydrocephalus. In the author's opinion, most small breed dogs normally have large lateral ventricles as a breed characteristic (ventriculomegaly) and are not hydrocephalic. In the absence of concurrent disease processes, CSF analysis is usually normal; occasionally, a mild mononuclear pleocytosis will be apparent, however.
Treatment of COMS
Treatment of COMS can be divided into medical and surgical therapy. In people with symptomatic Chiari type I malformation, surgical therapy is considered the treatment of choice, with foramen magnum decompression (FMD) being the preferred surgical procedure. Adjunctive surgical procedures are occasionally performed in people who have had a suboptimal response to FMD; such procedures usually involve placement of a shunt to divert syringomyelia fluid from the spinal cord region to another location for absorption (e.g., pleural or peritoneal cavity, subarachnoid space). Although there is a high degree of success in surgical management of Chiari type I malformation in people, there is a re-operative rate varying from 8%-30% for FMD; the most common problem necessitating re-operation is excessive scar tissue formation at the FMD site causing compression at the cervicomedullary junction, effectively recreating the original disease state. Medical therapy for dogs with COMS generally falls into three categories: analgesic drugs (implies relief of dysesthesia/paresthesia also), drugs that decrease CSF production, and corticosteroid therapy. By far the most useful drug available for relief of scratching activity associated with syringomyelia is gabapentin (10 mg/kg body weight PO, q 8 hrs). It has been shown that neuropathic pain is accentuated over time due to up-regulation of the α2δ-1 subunit of voltage-gated calcium channels in dorsal root ganglion neurons and dorsal horn nociceptive neurons of the spinal cord. Gabapentin, and the newer gabapentin analog, pregabalin, are believed to exert their antinociceptive effects by selectively binding to the α2δ-1 subunit and inhibiting calcium influx in these neurons. Side effects of gabapentin are minimal, usually restricted to mild sedation, pelvic limb ataxia, and weight gain. There is no information available concerning the use of pregabalin in dogs. A dose of 2-4 mg/kg body weight PO, q 8hrs has been suggested, based on preliminary pharmacokinetic data in normal dogs collected by the author and colleagues. It appears that some dogs will maintain high plasma pregabalin concentrations with every 12 hr dosing as well. Whether pregabalin is more effective or has fewer side effects compared with gabapentin is unknown. Based on data collected from epileptic dogs, pregabalin side effects appear to be similar to those reported for gabapentin. Orally administered opiate drugs are sometimes helpful in alleviating neck and head pain in COMS dogs. The author has had success using oral tramadol (2-4 mg/kg, q 8-12 hrs). A number of drugs aimed at decreasing CSF production have been used in COMS patients, in an effort to diminish the CSF pulse pressure. All information regarding efficacy of these drugs is anecdotal. They include omeprazole (a proton pump inhibitor), acetazolamide (a carbonic anhydrase inhibitor), and furosemide (a loop diuretic). More specific information regarding these drugs is covered in the congenital hydrocephalus discussion. Corticosteroids are often used in medical management of COMS. Potential benefits include anti-inflammatory effects, decreased CSF production, and decreased substance P (a nociceptive neurotransmitter) expression in spinal cord dorsal horn neurons. An initial antiinflammatory dose of 0.5 mg/kg PO, q 12 hrs is often effective in controlling clinical signs. This dose should be tapered, if at all possible, to an every other day schedule within the first month of therapy. In most cases of COMS, medical therapy will diminish the severity of clinical signs, but resolution is unlikely. The preferred surgical procedure for treatment of COMS in dogs is FMD. Based on two similar reports, short term surgical success rates with FMD in dogs with COMS are approximately 80%. One report found an inverse relationship between the length of time clinical signs were present prior to surgical intervention and the extent of post-operative improvement. Unfortunately, there appears to be a disease relapse rate ranging from 25% to 47% of cases; most of these relapses are suspected to be due to excessive post-operative scar tissue formation at the FMD site. In most cases, clinical signs of pain are routinely relieved with surgery, but scratching activity tends to persist. Recently, the author has adapted a cranioplasty procedure used in human FMD surgery to discourage excessive post-operative scar tissue from recompressing the operative site. Although results are preliminary, the cranioplasty procedure appears to have substantially decreased the re-operation rate for this disorder in dogs. There is little information regarding the prognosis for COMS in dogs. Most dogs with COMS will respond favorably to medical therapy, although in many cases this response is temporary. In one group of 10 COMS dogs treated medically, 5 dogs were euthanized within 2 yrs due to disease progression and diminished responsiveness to therapy. In another study, 36% of COMS dogs treated medically were euthanized due to clinical signs of their disease at mean of 1.7 yrs from the time of diagnosis. Although the surgical success rate is generally favorable for COMS in dogs, the recurrence rate due to excessive post-operative scar tissue formation is unacceptably high. Hopefully, refinements in surgical technique, such as cranioplasty, will ameliorate this problem. In general, the overall prognosis for COMS in dogs is guarded to good for sustained improvement in clinical signs.
1. Cerda-Gonzalez S, et al. J Vet Intern Med 2006; 20: 736.
2. Cerda-Gonzalez S, et al. J Vet Intern Med 2006; 20: 737.
3. Dewey CW, et al. J Am Vet Med Assoc 2005; 227: 1270.
4. Dewey CW, et al. Vet Surg 2007; 36: 405.
5. Guay RD. J Geriatr Pharmacother 2005; 3: 274.
6. Levine DN. J Neurol Sci 2004; 220:3.
7. Rusbridge C. Vet Surg 2007; 36: 396.
8. Rusbridge C, et al. J Small Anim Pract 2007; 48: 432.
9. Rusbridge C, et al. J Vet Intern Med 2006; 20:469.
10. Rusbridge C, et al. J Vet Intern Med 2004; 18: 673.