ProposalIn the absence of a diagnostic test specific for multiple sclerosis (MS), the neurological community has adopted diagnostic criteria for MS. The Schumacher criteria 1 were followed by the Poser criteria, 2 which were replaced in 2000 by what have become known as the McDonald criteria. 3 The criteria reaffirmed the necessity to demonstrate dissemination of clinical events and lesions in space and time and, importantly, to exclude other reasons or diseases explaining the neurological deficits and symptoms. In the McDonald criteria, it was acknowledged that 'MRI [magnetic resonance imaging], analysis of cerebrospinal fluid (CSF), and visual evoked potentials (VEP), can add to a clinical diagnosis' but that each has limitations of sensitivity and specificity 3 . However, with each successive revision of the McDonald criteria, 4,5 MRI has gained and CSF has lost in importance.With the availability of disease-modifying drugs, early diagnosis has become increasingly important and neurologists are exploring ways of diagnosing not only established or clinically definite (CD) MS but also clinically isolated syndrome (CIS) with high risk for conversion to MS, and even radiologically isolated syndrome (RIS) which can perhaps be thought of as a diagnosis in search of its patient. Admittedly, MRI can show dissemination in space and time which CSF cannot. However, MRI is often inconclusive, especially at early stages of the disease. CSF analysis adds a different type of information -inflammation and immunological disturbance. In addition, one important aspect of CSF analysis is to exclude other diagnoses. We discuss why analysis of CSF still is important and should be resuscitated within the diagnostic criteria of MS.
Diagnostic valueKabat et al. demonstrated in 1942 that 'gamma globulin' was significantly increased in CSF from patients with MS and subsequently demonstrated that 'restricted heterogeneity' or oligoclonal bands (OB) on electrophoresis were present in CSF but not in serum. 6 Since then many publications have shown the diagnostic value of CSF analysis and also, that in certain clinical situations, it can have predictive value.In patients with CDMS an elevated IgG-index is present in 70-80% and OB are demonstrated in CSF but not serum in almost every patient, provided the analysis is carried out in a good laboratory with the recommended method of isoelectric focusing. 7 Recent estimates in the Nordic countries suggest that around 90-95% of patients with CDMS have OB. 8,9 It is clear that the frequency of OB varies between countries and populations, and averages 80% globally. 10 One trivial reason for these differences is the quality of assays performed. However, other factors may also be in play. For example, the MS-predisposing DRB1*15:01 class II gene variant is correlated with a high frequency of OB, while DRB1*04 is correlated with OB negative CSF, 8,9 and the frequency of these class II gene variants differs considerably between populations. Still, if present and if other causes of OB are excluded such...