Syringomyelia, diagnosis and treatment
af Jörg Klekamp og Madjid Samii


Chapter 4

Syringomyelia associated with diseases of the spinal canal.

This chapter will describe the management of patients with disorders of the spinal canal that have caused a syrinx. A major problem in this group may be the exact localization of this underlying pathology. Tumors can be easily visualized with gadolinium enhanced MRI. All other entities described below may offer diagnostic problems, so that counseling of the patient, indication for surgery, and planning the surgical strategy may be quite difficult. Nevertheless, it is our firm belief that idiopathic syringomyelia does not exist. In each patient, it is possible to identify the cause of syrinx formation. Therefore, every effort should be made to demonstrate the underlying disorder.

Basically, the evaluation and treatment of these patients follow the same guidelines as outlined for patients with craniocervical disorders. Once a tumor has been excluded, clinical and neuroradiological evaluations have to be performed to demonstrate the exact extent and morphology of the syrinx and the presence of pathologies that may interfere with CSF circulation and / or spinal cord mobility, i.e., a tethered cord. Treatment should primarily aim at this underlying pathology. This chapter will describe how patients with syringomyelia related to spinal diseases should approached in terms of diagnosis and management.

Syringomyelia related to spine arachnoid scarring

The pathophysiology of syringomyelia related to trauma or arachnoiditis is still incompletely understood. Hematomyelia related to spinal cord trauma and spinal cord ischemia due to arachnoiditis are generally believed to cause syringomyelia in these instances. However, hematomyelia and cord ischemia will lead to instant neurological problems, i.e., myelomalacia, and not to a slowly progressing syrinx after years of delay. Furthermore, the development of a syrinx after mild spinal trauma or a small area of arachnoiditis cannot be explained by either ischemia or hematomyelia of the spinal cord.

The relevance of arachnoid changes for changes of CSF flow, spinal cord mobility, and the development of syringomyelia in these instances is not widely recognized. Shunting of the syrinx to either the subarachnoid, pleural, or peritoneal space is still widely accepted as the treatment of choice for patients with syringomyelia, due to trauma or arachnoiditis.

It has been known for a long time that spinal arachnoiditis or arachnoid scarring may be associated with syringomyelia. Many authors, for instance, report on arachnoid adhesions at the level of the spinal injury in patients with posttraumatic syringomyelia. However, the relevance of arachnoid scarring for neurological symptoms related to spinal cord tethering and syringomyelia has only recently become more appreciated.

Arachnoid scarring may be caused by a variety of mechanisms. Inflammatory reactions of the arachnoid due to bacterial, fungal or viral infections chemical substances such as contrast media or breakdown products of blood after subarachnoid hemorrhages or due to mechanical irritation related to trauma, degenerative diseases of the spine, scoliosis, kyphotic deformities or instabilities can lead to arachnoid scarring in a focal, circumscribed area of the spine canal. Another important cause of arachnoid scarring is intradural surgery. Any surgical intervention in the spinal subarachnoid space may lead to disturbances of CSF flow and even spinal cord tethering at the site of surgery. Depending on the cause of arachnoid pathology, the scarring may be limited to a very circumscribed area of just a few millimeters or may extend over several spinal segments. The most severe forms of spinal arachnoiditis are reported after tuberculous meningitis.

Even with a carefully documented patient history, the cause of spinal arachnoid scarring cannot be clarified in each individual patient. The most likely explanation in such patients may still be posttraumatic arachnoid scarring due to small tears or hemorrhages from little blood vessels inside the arachnoid layer without any neurological symptoms at the time of trauma.

For the purpose of this book, we suggest the terms posttraumatic syringomyelia (PST) for patients with spinal arachnoid scarring and a history of spinal trauma at that level and postinflammatory syringomyelia (PIS) for patients with spinal arachnoid scarring but no such history.

Clinical Presentation

Our series of patients with spinal arachnoid pathologies consists of 154 patients. In 68 patients, arachnoid scarring resulted from trauma, whereas 86 patients developed scarring as a result of an inflammatory reaction. Except for differences in sex distribution, there was no statistical difference between both groups of patients in terms of age, history, or follow-up.

Posttraumatic syringomyelia

Posttraumatic syringomyelia requires a traumatic lesion of the arachnoid. It does not necessite spinal cord trauma. Tethering of the spinal cord and CSF flow obstruction due to arachnoid scarring or spinal canal stenosis may the cause progressive syringomyelia. However, severe forms of spinal trauma will lead to a combination of soft tissue, bony, leptomeningeal and spinal cord lesions in various combinations. If spinal cord tissue is damaged severely, a necrotic area may result in a defect inside the spinal cord , i.e., myelomalacia. As a general rule, the extent cord of myelomalacia is restricted to the spinal level of the cord injury and does not progress. PST, however, begins at the level of the injury and may progress from there in either direction. Clinically, myelomalacia is almost always associated with instant symptoms related to the extent of spinal cord damage but without further clinical progression.

In syringomyelia, progressive clinical symptoms start to develop after an interval of typically several years. The neurological distribution of these new symptoms are related to the spinal representation of the syrinx. Especially after severe spinal cord trauma, myelomalacia and syringomyelia may coexist. Therefore, the term posttraumatic cystic myelopathy was introduced which may be separated into a atationary, i.e., myelomalacia, and a progressive i.e., syringomyelia, form.

The first description of a posttraumatic cystic lesion of the spinal cord was given by Bastian in 1867. In 1898 Cushing described that two patients with hematomyelia and supposed that residual symptoms of these patients were related to cyst formation after resorption of the hematoma. Lloyd postulated as early as 1894 a pathophysiological relationship between posttraumatic arachnoid scarring and cyst formation. Collier explained posttraumatic arachnopathy on the basis of small hemorrhages into the arachnoid or the subarachnoid space in 1916.

The incidence of PST is not precisely known. About 2% of patients with spinal cord trauma will develop new neurological symptoms with each ongoing year for a number of possible reasons, such as spinal instability, spinal cord compression or syringomyelia. Before the era of MRI, the incidence of PST was assumed according to clinical and autopsy data somewhere between 1.25% and 8%. Whit routine MRI examinations of patients with spinal cord trauma, Backe et al. gave a figure of 51% for patients with posttraumatic spinal cord cysts after an average interval of 1,6 years. Similar figures were given be several other authors. However, no distinction was made between syringomyelia and myelomalacia and patients with spinal trauma without affection of the spinal cord did not enter these studies.

There is no clear-cut relationship between spinal level of the trauma and the likelihood for PTS to develop. Several studies described syringomyelia to be more conmen after thoracia injuries, whereas others found syringomyelia more often associated with a cervical trauma. Discrepancies between older and more recent studies on this point may be explained by a reduced mortality rate of patients after cervical injuries.

We have seen 68 patients with PTS at an average age of 45 +/- 13 years. Males predominated by a factor of 2,9:1. 36% developed a syrinx after a cervical trauma, 53% after thoracia, and 11% after conus medullaris injuries. Compared with postinflammatory arachnoid scarring, there was a significantly higher percentage of patients with cervical scarring an fewer thoracic cases.

Seventeen patients were not operated on as they did not demonstrate progressive neurological symptoms. The remaining 51 patients underwent 76 operations. The mean interval between trauma and development of symptomatic syringomyelia was extremely variable. For one patient, it took 38 years before the first clinical signs of a posttraumatic syrinx were observed. Another patient showed signs of progressive neurological deterioration 2 weeks after the injury. Typically, several years (mean 9 years and 9 months) elapsed before syrinx became symptomatic. Some authors described a shorter interval after complete spinal cord injuries whereas others denies such correlation. After clinical symptoms of syringomyelia had started, it took another 51 +/- 70 months until patients presented to us.

A complete spinal cord injury preceded the syrinx in 24% of cases, an incomplete lesion in 40%, while 36% developed is similar to the observation by Barnett and Jousse who described PTS to occur more often after incomplete than complete lesions, whereas other authors made the opposite observation. Most authors, however, could not find a correlation between severity of trauma an occurrence of PTS. Posttraumatic arachnopathy is the key for the development of a syrinx and not the amount of cord damage due to the initial trauma. Several authors noted that a syrinx may develop even without an initial cord injury. A leptomeningeal hemorrhage may be sufficient to cause scarring and syringomyelia. It has been suggested that bulging of the yellow ligiment toward the spinal cord may lead to such leptomeningeal injuries.

PTS started in most instances with progressive motor weakness (30% of patients) or pain (28% of patients). Gait problem (12%) and dysesthesias (16%) were encountered less often. Sensory problems (9%), sphincter disturbances (2%), swallowing problems (1%), or cardiorespiratory dysfunctions (1%) were rare initial presentations. 51 patients demonstrated clinical signs of neurological progression. The major complaint at the time of surgery was local pain or pain radiating into dermatomes representing the spinal levels of the syrinx for 36% of patients and gait problems for 23%. Dysesthesias (18%) and motor weakness (15%) were less often considered to be the main clinical problems. Sensory deficits (4%), sphincter disturbances (1%), swallowing problems (1%), or cardiorespiratory dysfunctions (1%) rarely predominated the clinical picture. In the literature, pain and sensory deficits are usually considered to be the most common symptoms, whereas motor weakness tends to develop at a later stage.

Postinflammatory syringomyelia

For 86 patients, spinal arachnoid pathology could not be related to a particular traumatic incident. The underlying process that had caused this inflammatory reaction could be identified in 34 patients, 9 patients provided a history of meningitis, 3 patients had suffered from spondylodiscitis or spinal abscess formation, 17 patients had undergone intradural surgical procedures, 3 peridural anesthesia and 2 patients had suffered a subarachnoid hemorrhage. For 52 patients, however, the cause of arachnoid scarring remained obscure, i.e. 60% of patients developed symptomatic syringomyelia without any preceding, clinical problems. One may speculate whether minor traumatic accidents, which the patients did not recall, might have been responsible. As even a minor trauma may cause arachnoid scarring and symptoms of PTS usually appear after several years, this seems to be the most likely explanation.

42 patients underwent 65 operations, whereas 44 patients were not operated on as they did not demonstrate clinical signs of neurological progression. For 16% the arachnoid pathology was located in the cervical spine, for 71% in the thoracic area, and for 12% arachnoid scarring was situated at the level of the conus medullaris. Compared with posttraumatic arachnoid scarring, this distribution was significantly different in favor of thoracic postinflammatory scarring.

The average interval between inflammatory reaction and onset of symptoms related to syringomyelia was just over 9 years and similar to observations for posttraumatic patients. In one patients, it took 50 years for the syrinx to develop and to produce clinical manifestations. Once symptoms had appeared, it took another 4 years on average before the diagnosis was made and the patient was sent to a neurosurgeon.

The most common symptoms at the beginning of clinical manifestations of syringomyelia were gait ataxia (28%), back pain radiating to dermatomes of the syrinx (23%), dysesthesias (21%), and motor weakness (16%). 42 patients demonstrated clinical progression of neurological symptoms so that surgery was indicated. During such a progressive course, however, the clinical picture only change in terms of severity, patients still reported those symptoms as their major complaint with which syringomyelia had started to be clinically noticed.

Apart from the fact that a significant number of patients with PTS develop syringomyelia on top of a complete or incomplete spinal cord lesion with instant neurological symptoms, the clinical course of patients with PIS or PTS is not very different. The worse preoperative neurological status of posttraumatic patients is related to the severity of spinal cord trauma and not to different dynamics of syrinx progression.

Kilde: Syringomyelia, Diagnosis and Treatment by Jörg Klekamp and Madjid Samii. Springer 2002.

TIL TOPPEN