Most cases of syringomyelia are associated with Chiari malformation type 1, which in turn comprises the commonest abnormality encountered at the craniovertebral junction. It is characterized by underdevelopment of the posterior cranial fossa with overcrowding of an otherwise normally developed hindbrain (Milhorat et al. 1999; Nishikawa et al. 1997). A ubiquitous feature is compression of the retrocerebellar CSF spaces, and about nine out of ten cases have a tonsillar herniation that is at least 5 mm below the level of foramen magnum. Very commonly there are also radiographic signs of cranial base dysplasia , of varying degree (Milhorat et al. 1999). Chiari malformation type 1 has a reported male to female ratio of between 1:0.7 and 1:3.7 (Da Silva et al. 2011; Meadows et al. 2000; Milhorat et al. 1999; Takeuchi et al. 2007). The reported rate of Chiari malformation type 1 as an incidental finding on MRI of the brain ranges from 0.04 to 0.9 % (Meadows et al. 2000; Morris et al. 2009; Vernooij et al. 2007). The reported incidence is higher from studies using high-resolution MRI sequences. One study reported cerebellar tonsillar herniation in as many as 14.4 % of patients presenting with neck pain and/or upper limb symptoms (Takeuchi et al. 2007).

The reported occurrence of syringomyelia in association with Chiari malformation type 1 ranges from 65 to 80 % (Speer et al. 2003). Chiari type 1-related syringomyelia has also been reported as an incidental finding on MRI (Meadows et al. 2000; Nishizawa et al. 2001).

Chiari malformation type 2 is found only in patients with myelomeningocele. It is the leading cause of death in affected individuals under the age of 2, and up to 15 % of patients with early clinical manifestation of Chiari malformation type 2 die by the age of 3 years and nearly a third of survivors have some form of permanent neurological disability (Stevenson 2004). Outcomes in older children, presenting with myelopathy and/or pain, are much better, ranging from 79 to 100 % improvement in symptoms following surgery. The prevalence of Chiari malformation type 2 in the general population is 1 in 3,600.

Chiari malformation type 2 is associated with syringomyelia in 35 % of cases (Speer et al. 2003), and it accounts for up to 8 % of the total cases of syringomyelia, with a higher percentage in paediatric practice.

Borderline tonsillar herniation , 2–4 mm below the foramen magnum, has an estimated prevalence of 2.6 per 100,000 population, from all MRI scans of the brain (Takeuchi et al. 2007). Syringomyelia was found in just over half of these patients (Milhorat et al. 1999).

The definition of Chiari malformation type 1 is evolving from that of a simple anatomical description to the concept of it representing the clinical expression of a number of different pathologies. Five broad mechanisms causing cerebellar tonsillar have been described (Table 2.5) (Milhorat et al. 2010; De Souza et al. 2011). These include those which affect the development of the craniocervical structures. For example, the development of Chiari malformation type 1 was seen in 29 % of patients suffering from rickets and in 73 % of all cases of Crouzon ’s disease. A tight filum terminale has an associated Chiari malformation type 1 in 10 % of cases. Twenty-four percent of pseudotumour cerebri patients had inferiorly displaced cerebellar tonsils. In addition, venous sinus occlusion can be the cause of reversible hindbrain herniation (Novegno et al. 2008). The frequency of these different causes of cerebellar herniation is very variable (Table 2.6) and also has ethno-geographical differences (Da Silva et al. 2011; Milhorat et al. 1999, 2009, 2010; Novegno et al. 2008; Strahle et al. 2011a). Our own observation of 900 adult patients with Chiari malformations, over a 10-year period, found basilar impression in 17 % and non-syndromic craniosynostosis in 7.4 %. The incidence of hydrocephalus, when defined as an Evans’ index greater than 0.30, was present in as many as 54 % of patients. In contrast, the prevalence of basilar impression associated with Chiari malformation in the northeast of Brazil was more than 60 % (Da Silva et al. 2011). It may well be that differences in the frequency of cranial constriction, cranial settling and mild deformations of cranial shape can explain ethno-geographical differences in syringomyelia prevalence.

The causes of spinal cord injury vary from country to country, depending on social and economic factors. Post-traumatic syringomyelia was previously thought to be an infrequent but serious sequel to such injuries, and clinical and CT studies suggested that it occurred with an incidence of between 1 and 5 % (Barnett et al. 1971; Biyani and El Masry 1994; El Masry and Biyani 1996). Since the introduction of MRI, the reported radiological incidence has increased up to 22 % (Burt 2004; Squier and Lehr 1994), which is consistent with the frequency of 17–20 % identified in post-mortem studies (Squier and Lehr 1994; Wozniewicz et al. 1983). Cystic necrosis of the spinal cord, confined to the level of injury, is generally considered to be a myelomalacic cavity and not syringomyelia, but asymptomatic cavitations, extending above and below the levels of injury, are often detected radiologically in victims of spinal cord injury, and these outnumber cases of symptomatic post-traumatic syringomyelia. Which asymptomatic cavities are likely to become symptomatic and over what length of time is, however, unknown. Progression may depend upon the original mechanism of injury or a variety of conditions inherent to the individual or both (Byun et al. 2010; Ohtonari et al. 2009).

Males are more likely to be victims of spinal cord injury than are females, in a ratio of about 6:1 (Burt 2004; El Masry and Biyani 1996). The interval between injury and diagnosis ranges from 2 months to 34 years (Biyani and El Masry 1994; El Masry and Biyani 1996). Full neurological recovery following the original spinal cord injury does not eliminate the possibility of post-traumatic syringomyelia developing later.

Syringomyelia is a rare sequel (less than 1 %) of infectious and non-infectious central nervous system inflammatory disease (Williams 1995). There are two main mechanisms by which inflammation may lead to the formation of syringomyelia: arachnoiditis and myelitis . Infection may also be a factor precipitating the onset of symptoms in Chiari-associated syringomyelia, in up to 7 % of patients (Milhorat et al. 1999).

Primary spinal syringomyelia is commonly secondary to post-inflammatory scarring, which leads to obstruction to the normal spinal CSF flow. Arachnoiditis might also cause syrinx formation by causing obliteration of the spinal microvasculature, leading to local cord ischaemia. Patterns of arachnoiditis seen range from focal meningeal cicatrix formation to diffuse adhesive spinal arachnoiditis (Caplan et al. 1990).

Foramen magnum arachnoiditis, in the absence of Chiari malformations, is a rare cause of syringomyelia (Klekamp et al. 2002). The mean interval between the presumed causative event (meningitis or trauma) and the development of syringomyelia-related symptoms can be up to 10 years. Compared with patients with Chiari malformation type 1, individuals with syringomyelia due to foramen magnum arachnoiditis have a much poorer long-term outcome. A stable clinical course was demonstrated in only 14 % of patients in whom surgery was not performed. Following surgery, 57 % of patients will have recurrence of symptoms within 5 years of the procedure.