Magnetic resonance imaging was first used in 1981 to study patients with MS. The technique was rapidly accepted in the clinical realm. Images are generated as a result of energy release from protons (positively charged hydrogen nuclei) that have been aligned in the axis of a strong magnet. Water, lipids, and other molecules contain hydrogen atoms, which spin and precess around the main axis. This axis is referred to a magnetization vector. Radio frequency pulses cause the protons to rotate and spin away from the main axis. Once the radio frequency pulse is removed, a signal is induced as the protons return to their original rotation. The signal decays over time and is referred to a longitudinal decay (T1) and transverse relaxation (T2) time. T1 and T2 times are the result of the environment in which these proteins exist (e.g., CSF protons require a longer time to return to equilibrium). The radio frequency pulses can be given at different times and for different lengths of time to generate different types of images that are "weighted" toward T1 or T2. This weighting is based on repetition time (TR) or time between pulses and echo time (TE), the time to generate the echo. T1 images have both short TE and TR, whereas T2 images have both long TE and TR. CSF is white in T2-weighted images, as are demyeli-nating lesions of MS (Fig. 2). On T1-weighted images some chronic lesions appear as dark areas or "black holes,'' which indicate tissue destruction and probable axonal loss. Newer techniques, such as fluid-attenuated inversion recovery imaging, cancel the signal from CSF, and as a result demyelinating periventricular lesions are more easily detected (Fig. 3). Gadolinium, a rare element with paramagnetic substance, is given intravenously to provide contrast enhancement of areas of breakdown of the blood-brain barrier. Because gadolinium shortens the T1 and T2 relaxation time, images appear bright where the blood-brain barrier is disrupted. This enhancement, which is the earliest standard MRI evidence of a newly developing MS lesion, indicates acute inflammation and often persists 2-6 weeks in areas of demyelination.
In general, the lesions of MS are ovoid or round and located adjacent to the body or temporal horn of the ventricles, in the corpus callosum or infratentorial or cortical-subcortical areas. Size ranges from a few millimeters to larger confluent areas. Lesions change in size over time and may disappear, however, often an abnormality will persist on T2-weighted images. The white matter may appear diffusely abnormal (i.e., "dirty white matter''). Cortical lesions are present pathologically but more difficult to detect with MRI. New lesions or enlargement of old lesions occur while other lesions are shrinking. Spinal cord lesions are detected most often in the cervical region. The spinal cord may appear swollen during the acute phase and may be enhanced with use of gadolinium. Optic nerve abnormalities are visualized with special techniques to suppress signal from orbital fat.
Because of the frequency of nonspecific white matter changes in MRI, various criteria have been suggested to define when brain MRI abnormalities are consistent with the clinical diagnosis of MS. One frequently used set of criteria defines significant findings as three or more lesions, that are equal to or greater than 3 mm, with at least one located in periventricular or infra-tentorial areas. At least 90% ofpatients with clinically definite MS have MRI abnormalities. Other explanations for white matter abnormalities exist and include patchy white matter abnormalities related to hypertension or diabetes mellitus. Lacunar lesions due to strokes appear isointense to CSF because of complete tissue destruction, which is not typical of MS lesions. Other individuals with risk for cerebrovascular disease
may have incidental lesions scattered in the deep white matter. Lesions seen with systemic lupus erythemato-sus are also located in the deep white matter but are not typically periventricular. Neurosarcoidosis may cause white matter abnormalities but frequently has associated meningeal enhancement due to leptomeningitis. ADEM may appear identical to MS. Lesions tend to resolve and do not recur. Gliomas may appear as solitary lesions, as may "pseudotumor" lesions of MS. Follow-up MRI may be definitive, although biopsy may be necessary to distinguish the two. Finally, inherited disorders of myelin, the leukodystrophies, demonstrate symmetric confluent rather than patchy, discrete lesions. A clinical history and examination are most helpful in providing alternate explanations for these abnormalities.
MRI has provided much information about the underlying pathophysiology of MS. Serial imaging has demonstrated that white matter lesions are present even when individuals are not clinically aware of disease activity. This is not a consistent process and varies from month to month, however, the frequency of subclinical lesions may be 5-10 times greater than clinical disease. Clinically evident lesions tend to be present in the brain stem and spinal cord. Subclinical lesions in the cerebrum tend to increase with clinical events. The extent of lesions present on T2 scan does not correlate well with the clinical examination, other than a possible correlation with cognitive dysfunction. However, an increase in the extent of lesions over time correlates with increasing physical disability. The number and area of contrast-enhancing lesions does not correlate with disability. In primary progressive MS, there are often very few lesions present on T2-weighted images, compared to relapsing-remitting or secondary progressive MS in which multiple new and
enhancing lesions tend to develop over time. This may indicate a pathophysiologic difference in disease type. MRI also has a predictive value for the diagnosis of MS in individuals who have had monosymptomatic disease, such as transverse myelitis or optic neuritis. The risk for an eventual diagnosis of clinically definite MS is higher in those individuals with abnormal brain MRI. MRI has been used as a tool in the clinical trial realm, allowing for a shorter clinical trial because it is a surrogate marker for disease activity. This becomes increasingly important as the partially effective therapies available will need to be incorporated into clinical trials of new agents, thus requiring more patients in order to prove efficacy. Problematic in the use of MRI has been clinical trials showing an effect on the clinical or MRI aspect but not vice versa. This apparent mismatch of data will need to be clarified.
Newer techniques add to our understanding of why T2-weighted images may not correlate well with clinical disability. Magnetization transfer imaging (MTI) relies on a different relaxation time of protons bound to macromolecules versus that ofthose protons that are freely moving in water. The normal appearing white matter of T2 weighted scans is abnormal when assessed with MTI, implying a more diffuse process than demonstrated by conventional T2 images. Magnetic resonance spectroscopy (MRS) uses MRI to generate a spectra of hydrogen or high energy phosphorus-containing metabolites. The spectra is used to define pathology of lesions, including tumors and demyelination. The spectra reveal major resonances from choline, creatine and phosphocreatine, and N-acetylaspartate (NAA). Choline is present in membranes and also present with increased myelin breakdown products. Creatine and phosphocreatine tend to be stable in MS other than acute lesion. NAA is present only in neurons and neuronal processes and can be used as a specific axonal marker in white matter. Decreased NAA is indicative of axonal pathology. MRS has shown a decreased NAA concentration throughout the brain of individuals with MS, implying diffuse injury beyond the site of lesions detected by routine imaging techniques.
Brain atrophy documented by MRI shows volume loss around the third and lateral ventricles and decreased corpus callosum area and brain width. Atrophy has been documented early in the course of the disease, before significant clinical disability, and is likely another indicator of underlying axonal damage.
Thus, standard and newer MRI techniques have shown that MS can be a subclinical process resulting in increasing MRI burden of disease. Findings also support the hypothesis that axonal injury occurs early in the course of the disease and may be more diffuse than previously thought.
Was this article helpful?