Inflammation of the Spinal Cord

Epidemiology

MS most commonly presents between ages 30 and 40 years but can occur at almost any age. Females are affected twice as often as males. The incidence of disease varies from 1 per 100,000 in equatorial regions to 30 to 80 per 100,000 in northern Europe and the United States.

Clinical Presentation

The clinical presentation of spinal MS depends on lesion location. The most common symptoms are paresthesia, weakness, gait disturbance, and bowel and bladder dysfunction. Among patients who present with spinal cord lesions, 90% will have brain lesions visible on MRI. The natural history and prognosis of the various clinical forms of MS are presented elsewhere.

Pathophysiology

MS is a chronic, inflammatory demyelinating disease of the CNS. The pathogenesis is incompletely understood. The traditional understanding of the disease process has been that T cells autoreactive to myelin cause an inflammatory reaction, which is followed by recruitment of macrophages and microglia and myelin destruction. More recent studies have highlighted the heterogeneity of MS lesions and the heterogeneity of immunologic mechanisms, which vary with the stage of disease and from lesion to lesion at each stage. Alternative immunologic mechanisms of injury, including antibody and complement activation, hypoxic injury, and oligodendroglial metabolic defects, seem to play a role in many lesions. The apparent climatic variation in incidence has stimulated the hypothesis that pathogenesis may be, at least in part, related to infectious agents.

Genetic factors in MS are not well understood. There is a weak familial tendency, and those of western European ethnicity are at increased risk.

Pathology

MS lesions (plaques) show inflammation composed of T cells and macrophages. There is focal demyelination. Axons are usually at least partly preserved.

Imaging

MRI is the imaging modality of choice. All patients with suspected spinal cord MS should have an MRI unless it is contraindicated. If the patient with spinal cord symptoms has a contraindication to MRI, myelography with post-myelogram CT should be performed to exclude a cord-compressing lesion as the cause of the symptoms. If patients present with spinal cord symptoms, and a lesion which may be MS is seen on spinal MRI, brain MRI should be performed, because 90% of patients will have concomitant brain lesions.

MRI

The typical spinal cord lesions of MS show high signal intensity on T2-weighted (T2W) images. They are peripherally located within the cord on axial images and occupy less than half of the cross-sectional area of the spinal cord ( Fig. 21-1 ). Typically, the lesions are less than two vertebral bodies in vertical length. The extent of T2 signal correlates with the degree of demyelination. On T1-weighted (T1W) images, lesions are isointense to hypointense. There is variable enhancement after intravenous contrast agent administration, especially in acute and subacute (1-2 months) lesions. Enhancement may be nodular, ring-like, or homogeneous. Unlike in the brain, fluid-attenuated inversion recovery (FLAIR) images are insensitive to lesions in the spinal cord. Cord edema and swelling are not prominent. Cord atrophy may be seen in the late stages.

Occasionally, MRI findings are indistinguishable from those of spinal cord tumor, with avid enhancement, edema, and mass effect.

It is challenging to try to apply advanced MRI techniques to lesions of the spinal cord, owing to the small size of the cord and artifacts from respiratory motion, cardiac motion, cerebrospinal fluid pulsations, and motion of the spinal cord itself. In patients with MS, diffusion tensor imaging of the spinal cord shows decreased fractional anisotropy and increased mean diffusivity in lesions and in normal-appearing white matter. Several authors report decreased magnetization transfer ratios in MS lesions and in the normal-appearing white matter of the spine.

Epidemiology

ATM has an incidence of about 4 new cases per million people per year in the United States. All ages are affected, but there is a bimodal age distribution with one peak between 10 and 19 years of age and a second peak between 30 and 39 years of age. There is no sex predilection.

Clinical Presentation

Affected patients present with acute or subacute signs and symptoms of motor, sensory, and autonomic dysfunction localizable to the spinal cord. There is usually a clearly defined rostral level of the sensory deficit. The symptoms progress to a maximum between 4 hours and 21 days, with bladder dysfunction, inability to move the legs, and paresthesia and numbness. Autonomic dysfunction is manifest as urinary urgency, inability to void, or bowel or bladder incontinence. Evidence of inflammation within the spinal cord can be provided by detecting pleocytosis or elevated IgG level within the cerebrospinal fluid, or by MRI that displays spinal cord enhancement after intravenous administration of a contrast agent. Patients with a spinal cord infarction or spinal vascular malformations may have similar clinical presentations and imaging appearances.

Pathophysiology

ATM may occur as an isolated “idiopathic” entity or in association with prior radiation, direct cord infection, systemic diseases (e.g., lupus, Behçet's disease, sarcoid), or nonspinal neoplasms (i.e., paraneoplastic syndrome). When the transverse myelitis is associated with a systemic disease, the pathogenesis varies with the systemic disease. For example, patients with lupus-associated transverse myelitis have CNS vasculitis. Neurosarcoid lesions are often associated with noncaseating granulomas (see later).

Paraneoplastic cases of transverse myelitis may be caused by autoantibodies formed against tumor antigens that are shared with or are similar to antigens found on neuronal cells. Such a pathogenesis can be found in some patients with small cell lung cancer who form an autoantibody called collapsin response mediator protein-5 (CRMP-5-IgG). Females with breast cancer may form an anti-amphiphysin IgG.

A variety of humoral and cellular immune derangements have been proposed as possible mechanisms for the tissue injury seen in idiopathic ATM. In about half the cases of idiopathic ATM there is an antecedent respiratory, gastrointestinal, or systemic illness. These cases are referred to as parainfectious to indicate that the cord injury may have been caused by direct infection of the spinal tissue by the organism, direct cord infection with cord injury caused by immune-mediated damage against the agent, or remote infection followed by a systemic immune response leading to an immune-mediated injury to the noninfected spinal cord. The list of infectious agents associated with idiopathic ATM is long and includes several herpesviruses, Listeria monocytogenes , Staphylococcus species, and Streptococcus . Idiopathic ATM has been associated with antigens against pinworms and even dust mites. ATM has also been reported after various vaccinations. In most of these reported associations, a direct causative effect has not been proven.

Optic neuritis associated with longitudinally extensive transverse myelitis (LETM) (covering more than three vertebral levels) may be designated as LETM or neuromyelitis optica (NMO). This entity is associated with the autoantibody NMO-IgG, which is directed against the cell membrane water channel protein aquaporin-4. This entity is discussed in more detail in the next section.

Some patients who present with idiopathic ATM have MS or will later meet the criteria for MS. In most cases, the spinal cord lesions of MS involve short segments (two or fewer vertebral body levels) and do not involve the cord bilaterally at the same level. Spinal cord involvement with MS was discussed earlier.

Many patients with idiopathic transverse myelitis have elevated levels of non–organ-specific serum autoantibodies such as antinuclear antibody (ANA), extractable nuclear antibody (ENA), or Sjögren's syndrome A antibodies (SSA). Whether these patients should be considered to have idiopathic transverse myelitis, and how they should be classified, is controversial.

There is no clear genetic defect or inheritance of ATM, although some of the diseases associated with myelitis have an increased incidence in association with certain HLA subtypes.

Pathology

Biopsy is not recommended in suspected cases of transverse myelitis. The histologic description of lesions is mostly anecdotal, with biopsies performed on tumor-like lesions or at autopsy. Principal findings include inflammation and variable demyelination. The inflammation may involve the gray and/or white matter. There is focal infiltration by monocytes and lymphocytes and astroglial and microglial activation. Later in the course of disease, macrophages are more prominent and there may be cystic areas.

Imaging

MRI is the imaging modality of choice for all patients with suspected ATM unless specifically contraindicated. If the patient with possible ATM has a contraindication to MRI, myelography with postmyelographic CT should be performed to exclude a cord-compressing lesion as the etiology of the patient's symptoms.

MRI

The MRI findings of ATM are nonspecific. MRI may even be normal during the first 24 to 48 hours of symptoms. Common findings include high T2 signal occupying more than two thirds of the cross-sectional area of the cord (88%) and extending over a length of three to four or more vertebral segments (53%) ( Fig. 21-2 ). Some cases (47%) showed a small area of signal isointense to cord in the center of the hyperintensity on axial T2W images. Spinal cord expansion (47%) and focal, peripheral cord enhancement (53%) after intravenous contrast agent administration may be seen. The thoracic cord is the most commonly affected region.

It is not always possible to distinguish between the imaging appearance of ATM and intramedullary spinal cord tumors. Generally, tumors enhance more strongly, may have a central cavity, and are less homogeneous on T2W images. Tumors more commonly expand the cord and/or deform the exterior contour of the cord. ATM is not usually associated with spinal cord cavity or syrinx, whereas tumors frequently are.

When trying to distinguish between ATM and the spinal cord lesions of MS, the longitudinal and crosssectional extent of the lesion is important. In most cases, the spinal cord lesions of MS are smaller and peripherally located in the cord. They do not involve the cord bilaterally and involve only short segments of the cord (two or fewer vertebral body levels). Lesions of ATM occupy more than two thirds of the cross-sectional cord area and extend three or more vertebral segments in length.

Evaluation of the spinal cord using diffusion tensor imaging (DTI) has been reported in patients with ATM. Renoux and associates found decreased fractional anisotropy in areas of T2 signal abnormality in patients with ATM, compared with the spinal cords of normal subjects. Interestingly, they also found areas of decreased fractional anisotropy in normal-appearing areas of the ATM patients' spinal cords, away from the lesions of signal abnormality evident on T2W imaging. In some cases, these fractional anisotropy lesions not visible on T2W images corresponded to the patient's symptoms. This implies that there are significant areas of spinal cord injury that may only be detectable when studied with DTI.