The diagnosis and management of multiple sclerosis

It will soon be 20 years since magnetic resonance imaging (MRI) became part of the diagnostic armamentarium of multiple sclerosis (MS), and almost 10 since immunomodulatory drugs became available for its treatment. These have been landmark events in the history of diagnosis and management of the disease. On sober reflection there are important questions regarding the benefits that have accrued to patients and physicians from these developments. For well over a century the diagnosis of MS was based on its traditional dissemination in space and time, supported by a characteristic constellation of signs and symptoms. Valuable diagnostic data, i.e. immunoglobulin G oligoclonal bands identified in the cerebrospinal fluid, and evoked potential studies, especially of visual stimuli, were added to the diagnostic process. The use of computer assisted tomography (CT) faded quickly because of the overwhelming superiority of MRI. Despite their high cost, these machines became available in hospitals and radiological laboratories with amazing rapidity, in particular in US. Within a very few years there were more MRI machines in Boston than in the entire UK. It is now generally but quite wrongly believed that MRI constitutes the essential basis for the diagnosis. It quickly became apparent that MRI was an extraordinarily useful tool in demonstrating lesions of brain and spinal cord, and in 1981 Young et al. published the first MR image of MS plaques (1). It was thus not unexpected that the procedure would be included when McDonald et al. published new criteria in 2001 (2). Although McDonald et al. discuss the use of MRI, the very first line of the scheme reads: Clinical presentation: two or more attacks; objective clinical evidence of two or more lesions. Additional data needed for MS diagnosis: none. A table lists MRI criteria for brain abnormality derived from the publication of Barkhof et al. (3) They are purely quantitative, lacking any kind of qualitative descriptive features. These are also non-specific, since they can be found in a number of demyelinating diseases that are different from MS (4–7). Indeed it is curious that they were chosen to be included in what was bound to become widely used diagnostic criteria, as they were not based on MRIs obtained from a series of clinically, reliably diagnosed cases of MS, but rather from the retrospective examination of patients who had had a second episode following what was called a clinically isolated syndrome (CIS) suggestive of MS ‘‘. In the text, but not featured in one of the tables, McDonald et al. (2) made an extremely valuable comment regarding spinal cord lesions, noting that they should be unequivocally hyperintense on T2-weighted images, be at least 3 mm but under two vertebral segments in length and occupy only part of the cross section of the cord in order to be considered as evidence of MS. That section has generally been ignored by radiologists and neurologists alike, Poser CM. The diagnosis and management of multiple sclerosis. Acta Neurol Scand 2005: 112: 199–201 Blackwell Munksgaard 2005. C. M. Poser