Toward an Understanding of Syringomyelia: MR Imaging of CSF Flow and Neuraxis Motion

Pathophysiologic mechanisms causing syringomyelia among patients with the Chiari I malformation have attracted considerable interest, but remain poorly understood. It is generally thought that alterations in CSF flow resulting from tonsillar obstruction at the foramen magnum lead to abnormalities in fluid pressures in the spinal canal and cord. But why do some patients with low tonsils develop syringomyelia and others do not? In this issue of the AJNR, Hoffman et al (page 151) attempt to elucidate the pathogenesis of Chiari-associated syringomyelia by measuring CSF flow patterns and cord pulsations. There are a number of technical limitations in this study, including the use of a low temporal sampling rate that may not accurately depict rapid changes in cord motion and CSF flow. Analysis of flow measurements was limited to comparison of maximum volumetric flow rates. CSF flow changes were observed, but were not statistically significant. Nonetheless, systolic and diastolic cord pulsations were higher among patients compared with healthy volunteers, and among patients with syringomyelia compared with patients without syringomyelia. This selective change in a dynamic parameter appears to correlate with disease and requires further evaluation.

The embryogenetic model of syringomyelia postulates pathologic alterations in neurogenesis as the basis of the structural abnormalities that lead to syringomyelia. The hydrodynamic theorists attempt to explain the changes in syringomyelia as the result of disturbances to processes that alter fluid dynamics. Alterations of flow parameters can thereby be interpreted as phenomena resulting from the underlying structural or hydrodynamic abnormalities. Conventional MR imaging provides the necessary anatomic information on structural parameters, such as degree of tonsillar herniation, amount of obstruction at the foramen magnum, distribution and size of the syrinx, as well as the presence of other anatomic lesions or tissue relaxation changes. Flow studies provide additional information about physiologic parameters, such as CSF flow patterns and motion of the neuraxis. Application of other techniques, such as diffusion imaging and spectroscopy, may eventually provide useful additional data on the intrinsic and metabolic properties of the neuraxis.

Presently, a number of studies using MR have reported changes in CSF flow, tonsillar motion, and syrinx characteristics. In certain cases, some parameters were found to vary with the degree of disease and the effects of operation. Undoubtedly these changes are related to the disease, but are measures primarily disease-specific markers or epiphenomena? The physiologic parameters that we presently observe are the effect, not the cause, of the disease—post hoc ergo propter hoc. Can the measurements aid in establishing a diagnosis and predicting disease formation, progression, and outcome? Unfortunately, it is presently unclear how changes in these structural and physiologic parameters are linked to the pathogenesis of syringomyelia. More specifically, the basic mechanisms involved in syringomyelia are still not well understood.

Although cord motion abnormalities may sometimes be associated with syringomyelia, they do not directly explain the pathogenesis. Caudal spinal cord motion is usually maximal at an instant of low CSF flow (near CSF end-diastole/beginning systole) and thereby does not appear directly related to systolic spinal pulse pressure abnormalities. In addition, the actual intraspinal fluid displacement caused by caudal cord displacement is very small, and should not cause significant pulse-pressure waves. Rather, the caudal motion of the cord, tonsils, and brain stem may be more significant in contributing to the resulting intraspinal (and intracerebral) pulse-pressure abnormalities through an increased obstruction at the foramen magnum. Present studies suggest that if the central canal does not communicate with the fourth ventricle (noncommunicating syringomyelia), exaggerated spinal pulse pressures may lead to increased transmedullary pressure gradients and force movement of interstitial fluid across the spinal cord. Depending on the degree of patency of the central canal and the nature of the intramedullary fluid pathways, syringomyelia could also develop in communicating and parenchymal syringomyelia; signal changes and reversible enlargement of the cord may develop in the presyrinx state.

Abnormalities in cord motion have been observed in a number of diseases. The hypothesis that increased cord motion may be associated with low tonsils and syrinx development is inconsistent with a number of pathologic observations. How would syringomyelia develop in Chiari patients where cord motion is decreased owing to tethering? Cord motion may not be increased among patients with low tonsils, with or without syringomyelia. In some cases, cord motion may actually be reduced because of scarring or significant compression at the foramen magnum. The appearance of low tonsils in intracranial hypotension can mimic Chiari and lead to misdiagnosis, but cord motion is decreased rather than increased. In fact, cord motion abnormalities in syringomyelia may not always be associated with low tonsils. Non-Chiari syringomyelia can be associated with etiologies such as arachnoiditis, soft-tissue lesions, trauma, and other structural spinal lesions in the subarachnoid space; it can also reveal abnormalities of CSF flow and neuraxis motion. The basic mechanism of syrinx development in these disorders may also involve increased CSF pulse pressures arising from obstruction of the subarachnoid space, similar to that found in Chiari-associated syringomyelia.

Although increased cord pulsations may reflect the degree of functional narrowing at the foramen magnum, they are not specific to Chiari-associated syringomyelia; they are seen in other entities with spinal canal narrowing. Surprisingly, cord motion may actually increase in cases with spinal canal stenosis, provided there is not any significant cord compression or tethering. In our experience, this phenomenon can occur in a number of different diseases involving partial restriction of cord mobility. Owing to an obligatory decrease in transverse cord mobility at the foramen magnum in Chiari, the intracranial systolic pulse pressure may be transmitted to the spinal cord primarily in a longitudinal direction, thereby increasing caudal cord motion. In addition to other pathophysiologic effects, such as altered cranial CSF ejection dynamics, the increased caudal cord impulse in Chiari may be a reflection of the decreased free space at the foramen magnum.

CSF flow abnormalities can be useful and sensitive measures for the evaluation and identification of a primary abnormality, but may originate at sites remote from the region of altered flow because of the complex nature of fluid wave propagation. In some cases, identifying the degree and location of neuraxis motion abnormalities may provide additional information to target the structural causes more accurately. Thus, observing both CSF and cord dynamics may be useful for identifying more precisely the nature of the offending pathology, especially in complex cases with multiple abnormalities.