Diagnostic Criteria for MS
Roberto Bomprezzi, MD––Barrow Neurological Institute, Phoenix, Arizona
Difficulties in Diagnosing MS
How can a physician know if a patient has multiple sclerosis (MS)? The kind of damage MS does to the nervous system is well known. MS is a progressive disease characterized by lesions in the central nervous system (CNS) in which myelin, the fatty material that insulates nerve fibers, is destroyed. Over time, most people with MS become increasingly disabled because of this damage. Despite this knowledge, however, no single test exists that can definitely identify MS, and making an accurate diagnosis early in the course of the disease can be quite difficult.
The crux of the MS diagnosis is showing that demyelinating lesions occur in several places in the CNS and that new lesions appear over time. This key concept is referred to as “dissemination in space and
time.” Evidence for dissemination in space and time can be drawn from the patient’s history, a clinical examination, laboratory tests, and imaging studies. These pieces of information are combined to determine the likelihood that the patient has MS. Over the decades, increasingly refined sets of criteria for diagnosing MS have been proposed. All of them focus on establishing these two diagnostic features that best define the characteristic presentation of MS.
At one time, there was no hurry to diagnose MS, primarily because little could be done besides helping the patient learn to cope with disabilities. In recent decades, however, research and drug development have changed this picture. Early treatment of MS has become important for two reasons. First, medications that can slow the progression of MS have been developed, and early treatment may signifi cantly slow or prevent CNS damage, in turn limiting the increased disability that accompanies progression of the disease. Second, it is now clear that MS not only destroys myelin, but also destroys nerve fibers. Although CNS cells can replace the damaged myelin over the short-term, the loss of nerve fibers is permanent and leads to the death of the neurons themselves. Moreover, the destruction of nerve fi bers begins early in the course of the disease. Given the urgency to begin treatment, the need for a rapid, reliable, and specifi c set of diagnostic criteria for MS has become ever more pressing.
The reason MS is so difficult to diagnose early on is because many symptoms are shared with other neurological illnesses. The most frequent symptoms of MS are altered sensations in the limbs that usually begin in one area and spread to involve the entire extremity. Also common are the loss of vision
in one eye, loss of coordination, and an abnormal gait. Many diseases exist that can mimic some or all of the symptoms of early MS. Chronic Lyme disease is a good example of a bacterial infection that can affect nervous system functions, and its symptoms can be confusingly similar to MS. Other bacterial
and viral infections can also cause MS-like symptoms. In fact, the infl ammation of part or all of the CNS, which sometimes follows an infection, can result in demyelination, a well-known feature of MS, and can manifest with the same impairment of function that MS would cause. Other disorders that can
mimic early MS include certain blood vessel diseases of the brain, some autoimmune conditions, and tumor growth. Clearly, these different diseases require different treatments, and some are life threatening in the absence of a quick, accurate diagnosis. Even in the presence of these confusing alternatives, the demonstration that demyelinating lesions are disseminated in space and time is a good indicator of MS, as long as the clinical picture is not better explained by different diagnoses.
Diagnostic Guidelines–First Attempts
With these issues in mind, scientists and physicians have repeatedly drawn up configurations of symptoms and test results that indicate a diagnosis of MS. The fi rst of these were the “Schumaker criteria,” which were published in 1965, and relied primarily on the results of a neurological examination
(modern laboratory tests and imaging systems were not available at that time). For the Schumaker criteria, dissemination in time was determined by the patient’s history, by the recurrence of exacerbations (periods of worsening symptoms, also called ‘relapses’), and/or progressively worsening symptoms. Dissemination in space relied on the neurologist’s ability to assign the symptoms to the portions of the nervous system likely to be responsible for them. For example, if a patient had visual problems during one exacerbation, but numbness of the legs during another, it would be clear that two different parts of the brain were involved. Dissemination in time required multiple exacerbations, or the worsening of symptoms over time, with or without relapses. Finally, the diagnosis of MS was made by exclusion when the symptoms could not be explained by a different diagnosis.
Between the mid-1960s and the late-1970s, along with methods to measure immune system status, imaging techniques were developed. By the early 1980s magnetic resonance imaging (MRI) was available for clinical use and the introduction of the MRI has literally revolutionized the practice of neurology. In particular, its application to the field of MS has allowed substantial advances in the understanding of the natural history of the disease. Besides MRI, other helpful tests for measuring immune reactions in the CNS were further refined. One that is still of great importance in the diagnostic evaluation for MS, consists of examining the cerebrospinal fluid (CSF), which bathes brain and spinal cord, for the presence of antibodies. Patients with MS often have antibodies in their CSF
that are not present in the blood serum, a finding that indicates that there is some sort of an immune response that is specific to the CNS and is suggestive of an abnormal process confined to that compartment. Another way to detect damage attributable to MS is to assess the function of the visual pathways, because MS-caused changes to the myelin of the optic nerves result in characteristic delays of the conduction of the impulses through those pathways. The test that measures that function is called VEP for “visual evoked potentials,” and it is used as an adjunctive tool to diagnose MS.
In 1983, Poser et al. proposed a new set of criteria for diagnosing MS that combined findings from the clinical examination and patient history with the results of MRI, CSF testing, and VEP. In the Poser criteria clinical symptoms alone were enough to diagnose MS if the patient had had two exacerbations and the examination showed two neurological abnormalities attributable to problems in different parts of the CNS; these clinical criteria thus established dissemination in space and time. In the absence of these clinical findings, the Poser criteria listed various combinations of exacerbations, clinical findings, and imaging or test results that could be used to establish the dissemination requirements. The Poser criteria were quickly adopted by clinicians and were used until quite recently.
Diagnostic Guidelines––The McDonald Criteria
The two decades after the Poser criteria were published were rich with studies that showed correlations between MS lesions and MRI findings, and medications that could slow the progression of the disease were developed. By the turn of the century, it was clear that enough had been learned that new diagnostic criteria were needed. In 2001, the International Panel on the Diagnosis of Multiple Sclerosis, headed by Dr. W. Ian McDonald, published a new set of criteria (see Tables 1 and 2), which became known as the “McDonald criteria”. These criteria still accepted a diagnosis of MS on the basis of clinical fi ndings and the patient’s history alone. For the fi rst time, however, the criteria listed combinations of imaging and test results that could be used to establish the MS diagnosis when the clinical findings were merely suspicious. The intent of the study group was to provide criteria that took into account both the need for an early diagnosis and the necessity of making an accurate diagnosis. For example, many people with MS present at the onset of their disease with visual symptoms (caused by optic neuritis, inflammation of the optic nerve), but not all people with such symptoms develop MS. The McDonald criteria could be used to determine if the visual symptoms were accompanied by brain lesions or other changes that strongly suggested that their cause was MS.
More importantly the McDonald criteria introduced the use of the MRI to establish the kind, number, and location of the demyelinating lesions that would be specific for a patient with MS. Extensive research on MS lesions has led to the recognition of patterns that were most commonly associated with an accurate MS diagnosis. The McDonald criteria provide guidelines that allow lesion patterns from individual patients with MS to be compared with those from other conditions and assess the likelihood of MS.
To understand the McDonald criteria, some additional information is needed. First, regarding MRI, two common types of MR sequences are always used “T2-weighted” and “Gadolinium (Gd)-enhanced.” In T2-weighted images, lesions containing extra fluid, indicating swelling and inflammation, show up as bright areas against a dark background. Gd is a contrast agent injected into the patient’s bloodstream. “Gdenhancing lesions” show up as very bright areas on the image. In these cases, Gd has leaked out of the bloodstream into the enhanced lesion, and this leakage indicates a breakdown in the normal separation of the bloodstream and the CNS. This happens as a new lesion is formed. Second, regarding the location of lesions, because MS lesions are found in areas where nerve fibers (but not neuron cell bodies) are densely packed, lesions that are most likely to represent MS are in these “white matter” areas.
In Table 1 “juxtacortical” refers to lesions in the white matter adjacent to gray matter (where the cell bodies are); “periventricular” refers to lesions near the ventricles, large spaces in the brain fi lled with CSF, around which many nerve fi bers course. The “posterior fossa” is where the nerve fibers exit the brain to descend through the brain stem and spinal cord. The 2001 McDonald criteria underscore the presence of lesions in these white matter areas as important pieces of imaging data that can lead to a diagnosis of MS. Because the McDonald criteria were very different from any that were published previously, the International Panel recommended that the criteria be tested carefully for accuracy. Shortly thereafter, the Consortium of MS Centers met and identified several areas in which the McDonald criteria could be improved (see http://www.mscare.org/cmsc/CMSC-Guidelines-on-MS.html). In addition, they recommended that clinicians use the McDonald criteria only as guidelines, mostly because some patients who would benefit from treatment at the time that the first demyelinating lesion was discovered would not fi t the criteria, and they did not want appropriate treatment to be withheld on the basis of the criteria.
In 2005, the International Panel reconvened to review the effectiveness of the criteria (Polman et al., 2005). They considered the information from a number of studies that had been performed during the previous 4 years, and used the data to make revisions. As shown in tables 1 and 2 , the changes were relatively minor. The MRI criteria for dissemination in space were altered to allow increased consideration of spinal cord lesions, because these can be helpful in distinguishing MS in difficult cases. The panel emphasized, however, that the nature of the spinal cord lesions had to be consistent with those typical for MS. For dissemination in time, the Panel recommended including new T2-weighted lesions appearing 1 month after disease onset, rather than relying on the presence of Gd-enhancing lesions occurring at least 3 months after onset, although the Gd-enhancing criterion was still considered applicable. This change allowed for a much quicker diagnosis. Finally, the Panel found that for the diagnosis of primary progressive MS (PPMS––MS that progresses from the time of onset, with or without superimposed relapses), a positive CSF test was not specific enough to be a requirement for diagnosis, as it had been in the 2001 criteria. In the 2005, criteria, the results from CSF antibody testing were to be used as supporting data, like the results from VEP.
In 2006, Swanton et al. published a study of patients who had Clinically Isolated Syndrome (CIS), in which a patient has a single incident of a neurological symptom similar to early MS, and objective evidence of a CNS abnormality (for example, from MRI, CSF, or VEP results). This is a good patient group for testing the validity of diagnostic criteria, because more than 80% of patients with CIS eventually develop “clinically definite” (CD) MS (that is, with time the MS symptoms become obvious enough to make a diagnosis on the basis of clinical findings alone). If the diagnostic criteria can predict which patients will develop CDMS, the criteria are established as sensitive and accurate. Furthermore, the amount of accuracy can be measured and compared for different criteria. In the 2006 study, the authors suggested simplifying the McDonald criteria to require one positive T2-weighted lesion in two of the four previously designated brain regions as proof of dissemination in space, and the presence of a new T2-positive lesion any time after a baseline image had been taken as proof of dissemination in time.
In 2007, the same group published a study in which they compared the ability of their simplified McDonald criteria to predict whether patients with CIS would develop MS with that of the 2001 and 2005 McDonald criteria. Because the new criteria used only MRI data (along with the clinical diagnosis
of CIS), only the criteria for MRI were compared. New guidelines were not suggested for PPMS. The study applied the three sets of criteria retrospectively to an analysis of 282 patients with CIS from MS centers in Barcelona, Amsterdam, Milan, and London. The two primary measures of accuracy for a diagnostic study are “sensitivity” (the ability of the test to pick out true positives) and “specifi city” (the ability of the test to identify true negatives). The 2007 study showed that the simplified criteria were more sensitive than either the 2001 or 2005 McDonald criteria, and had comparable specificity. The overall accuracy of the simplified criteria was better than for the McDonald criteria, and the usefulness of the 2006 criteria for predicting which patients would develop MS was similar to or better than the McDonald criteria for 2001 and 2005, respectively.
The authors also compared the accuracy of the criteria for dissemination in space separately from dissemination in time. For the distinct analyses of dissemination in space or time, the new criteria had better sensitivity but worse specifi city than the 2001 and 2005 McDonald criteria, but the overall accuracy was comparable for all three. The authors showed that all three sets of criteria were most accurate in predicting a patient’s conversion from CIS to CDMS, when a patient’s symptoms fulfilled the requirements for dissemination in space time, but that dissemination in time was more important for making an accurate prediction. Finally, although the authors believed that the data from their study indicated that their streamlined criteria could be used to identify MS more easily and as accurately as the McDonald 2001 and 2005 criteria, they emphasized that the streamlined criteria for dissemination in space could not be used accurately without considering dissemination in time.
Looking Ahead
The McDonald criteria and the Swanton group’s streamlined criteria provide clinically tested guidelines for making an accurate diagnosis of MS. Nonetheless, even quicker and more reliable criteria for diagnosing MS would clearly be welcome. The Swanton et al. (2007) study suggests that it may be possible to rely on MRI data as the supplement to clinical findings for an accurate diagnosis of MS. This possibility may be realized rather soon, because, as pointed out in both the 2005 revision to the McDonald criteria and the 2007 Swanton et al. study, more powerful MRI equipment is being developed, and these techniques are expected to reveal details of the CNS damage that are not detected by the MR imaging currently available. Once this new technology becomes clinically available, it may become possible to diagnose MS accurately from the MR images alone, and the diagnostic criteria will be further simplified yet achieving increased accuracy.
References
McDonald. W. I., Compston, A., Edan, G., Goodkin, D. E., Hartung, H. P., Lublin F. D., et al. (2001). Recommended Diagnostic Criteria for Multiple Sclerosis: Guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Annals of Neurology, 50, 121-127.
Picone, M. A., Cook, S. D. Multiple Sclerosis. Book Chapter, unpublished.
Polman, C. H., Reingold, S. C., Edan, G., Filippi, M., Hartung, H.-P., Kappos, L., et al. (2005). Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria.” Annals of Neurology, 2005, 58, 840-846.
Poser, C., Paty, D., Scheinberg, L., McDonald, W., Davis, F., Ebers, G., et al. (1983). New diagnostic criteria for multiple sclerosis: Guidelines for research protocols. Annals of Neurology, 13, 227-231.
Schumacher, G. A., Beebe, G. W., Kibler, R. F., Kurland, L. T., Kurtzke, J. F., McDowell, F., et al. (1965). Problems of experimental trials of therapy in multiple sclerosis: Report by the panel on the evaluation of experimental trials of therapy in multiple sclerosis. Annals of New York Academy of Sciences, 122, 552-568.
Swanton, J. K., Fernando, K., Dalton, C. M., Miszkiel, K. A., Thompson, A. J., Plant, G. T., et al. (2006). Modifi cation of MRI criteria for multiple sclerosis in patients with clinically isolated syndromes. Journal of Neurology, Neurosurgery, and Psychiatry, 77(7), 830-833.
Swanton, J. K., Rovira, A., Tintore, M., Altmann, D. R., Barkhof, F., Filippi, M., et al. (2007). MRI criteria for multiple sclerosis in patients presenting with clinically isolated syndromes: A multicentre retrospective study. The Lancet Neurology, 6, 677-686.