Relative value of three laboratory methods in the diagnosis of multiple sclerosis.

We compared three methods of analysis for IgG in cerebrospinal fluid, using samples from 158 patients with clinically suspected multiple sclerosis and from 200 neurological controls. The tests were: search for oligoclonal bands, calculation of rate of synthesis of IgG in the cerebrospinal fluid, and determination of the IgG/albumin ratio. Paired cerebrospinal fluid and serum samples were collected and their IgG and albumin concentrations measured. Oligoclonal bands were detected by electrophoresis on agarose. Positive results were obtained in 94, 75, and 67% of patients with probable or definite multiple sclerosis by the three respective methods. In contrast, for patients for whom the clinical diagnosis of multiple sclerosis was considered possible, positive results were obtained in 10, 43, and 13%, respectively. Evidently, detection of oligoclonal bands remains the best single test for the presence of abnormal IgG in suspected multiple sclerosis patients. A combination of the first two tests is most sensitive for both probable and definite multiple sclerosis (97%) and possible multiple sclerosis (50%). Some infectious or immunologic disorders can also produce these IgG abnormalities, but they can usually be distinguished from multiple sclerosis by other clinical and laboratory data.