CLIN. CHEM.26/1, 115-116(1979) CLINICAL CHEMISTRY, Vol. 26, No. 1, 1980 115 Immunoglobulins in Cerebrospinal Fluid in Various Neurologic Disorders A. Bouloukos, J. Lekakis, J. Michael, and A. Kalofoutis We determined the Concentrations of immunoglobulins A, G, and M in cerebrospinal fluid of 16 patients suffering from multiple sclerosis, 13 with non-bacterial meningitis, 10 with stroke syndrome, and 13 with epilepsy. The differences in concentrations of immunoglobulins in these groups were remarkable in the patients with multiple sclerosis, men- ingitis, or stroke syndrome. We propose that the deter- mination of the absolute immunoglobulin content in Ce- rebrospinal fluid is of greater significance than the relative immunoglobulin concentration. AddftlonalKeyphrases: multiple sclerosis nonbacterial meningitis . stroke syndrome #{149} epilepsy More than 35 years have passed since Kabat et al. (1), using the Tiselius electrophoresis method, first noted that the y-globulin content of cerebrospinal fluid (CSF) was increased in patients with multiple sclerosis and neurosyphilis. Since then, considerable research has been directed towards finding abnormalities of the CSF immunoglobulins G, A, and M (IgG, IgA, and 1gM), which are characteristic of various neurologic disorders, especially multiple sclerosis, and infective diseases of the central nervous system. Early in these studies Schneck and Glaman proposed (2) that the determination of the rel- ative concentration of immunoglobulins in CSF (i.e., expressed in percent of total CSF protein concentration) is of greater significance than the absolute immunoglobulin content (mg/L). The purpose of this paper is to determine the concentration of IgG, IgA, and 1gM in CSF in various established neurologic diseases and to discuss their possible role in the pathogenesis of these disorders. Materials and Methods Cerebrospinal fluid was sampled by lumbar puncture of 52 patients who were divided into four groups. Group I was 16 patients, 21 to 58 years old (mean, 33), suffering from multiple sclerosis diagnosed according to the following criteria: (a) age of onset between 15 and 45 years of age, (b) signs indicative of damage to at least two different parts of the central nervous system, and (c) a course of disease with at least two exacerbations and one remission. Group 2 was 13 patients, 6 to 42 years old (mean, 27), suf- fering from non-bacterial meningitis. Group 3 was 10 patients, 35 to 77 years old (mean, 48), with stroke syndrome. Group 4 was 13 patients, 22 to 53 years old (mean, 33), with idiopathic epilepsy (grand ma!). The control group consisted of 21 persons, 13 to 77 years old (mean, 34), suffering from minor psychoneurotic disorders, whose CSF leukocytes were less than 5/L and whose total CSF protein concentration was less than 450 mg/L. Cell count was routinely performed and the cells were separated by a gentle centrifugation; after the total protein was measured, specimens were stored at -20 #{176}C. We esti- Department of Biological Chemistry, University of Athens, School of Medicine, Goudi(609) -Athens, Greece. Received July 9, 1979; accepted Oct. 8, 1979. mated CSF immunoglobulins by the single radial immuno- diffusion method of Mancini et al. (3), using Behringwerke AG-Marburg/Lahn, 67019 Scoppito (l’Aquila) standard sera with plates covering the higher ranges. Statistical analysis was by Student’s t -test. Results The results of this study, given in Table 1, are summarized as follows: 1. Total CSF protein in the control group was 223±89 (SD) mg/L; in multiple sclerosis, it was 388 ± 179 mg/L, which is statistically significant. Also statistically significant was the increase in total CSF protein in non-bacterial meningitis. In stroke syndrome and epilepsy, however, the difference was not significant. 2. Absolute concentration of IgG in CSF of controls was 42 ± 21 mg/L, which differed significantly (p <0.025) from that obtained from multiple sclerosis patients. The increase in CSF IgG was also statistically significant in non-bacterial menin- gitis and in stroke syndrome, but not in epilepsy (60.6 ± 35 mg/L). 3. The relative concentration of IgG in CSF of healthy in- dividuals was not significantly different (p > 0.1) from that in patients with multiple sclerosis, non-bacterial meningitis, stroke syndrome, or epilepsy. 4. We detected IgA in only four of 21 cases in the control group (19%), but more often in the patient groups (present in 31 to 54% of the subjects; see Table 1). We could not detect 1gM in the controls or in the multiple sclerosis, stroke syn- drome, and epilepsy groups, and in only one case of non-bac- terial meningitis. Discussion Previous reports dealing with CSF immunoglobulins in various neurologic diseases have focused attention on the concentration of CSF-IgG in demyelinating and infective diseases of the central nervous system (1, 2, 4-7). Some in- vestigators suggest that the relative CSF-IgG concentration is of greater importance than the absolute CSF-IgG content (2). Our study does not support this suggestion: we did not find any differences in the relative CSF-IgG concentration, even in cases where the absolute CSF-IgG concentration was significantly increased. We were able to confirm previous observations of increased CSF-IgG in patients with multiple sclerosis (2,6-8), but we dispute the finding that the relative concentration of CSF-IgG is significantly increased in these patients. We believe that the increase in concentration of CSF-IgG is connected with the increase in total CSF protein; conse- quently, the ratio of IgG to total protein does not change sig- nificantly. This increase could be attributed to a migration of protein from serum to CSF because of damage to the blood-CSF barrier. The increased frequency of detection of CSF-IgA in mu!- tiple sclerosis (seven of 16 cases, or 44%) is interesting and agrees with previous observations (2, 7,9). This increase may be expected to occur in cases with damage to the blood-CSF barrier and increased total concentration of protein in CSF, CSF-IgA concentration increasing proportionally with the increase of total concentration of CSF protein (10). Further-