Materials and methods

PATIENTS

All patients with multiple sclerosis within the county of Hordaland, western Norway and with clinical onset within that county, during the period 1976 to 1986 and diagnosed before 1 January 1987, were selected for the study. The total of 144 patients comprised 61 men and 83 women, aged 17–66 years (mean age 39.2 years). According to the criteria of McAlpine,22 130 (90.3%) patients were classified as having definite multiple sclerosis, 10 (6.9%) as having probable multiple sclerosis, and four (2.8%) as having possible multiple sclerosis. Mean duration of disease was 6.9 years, range 1–11 years, and the initial course of the disease was relapsing-remitting in 117 (81.2%) and primary progressive in 27 (18.8%). Serum samples were collected in 1988 and 15 (10.3%) of the patients were in acute relapse at the time of blood sampling.

CONTROLS

Controls comprised 170 patients admitted to hospital with traumatic fractures, traumatic rupture of ligaments, or minor gynaecological or plastic surgery disorders. There were 77 men and 93 women, aged 18–77 years (mean age 40.0 years) and they were category matched to the cases according to age (five year age groups), sex, and area of residence (inland, coastal, and urban area) that gave a similar distribution of cases and controls for these variables. Blood sampling was performed at the same time as for the patients.

ANTIBODY ANALYSIS

IgG and IgM antibodies to EBV viral capsid antigen (VCA) were determined by enzyme linked immunosorbent assay (ELISA) (Dupont, N Billerica, MA, USA) and indirect immunofluorescence (IIF) (Organon Teknika, Durham, NC, USA). IgG antibodies to EBV nuclear antigen (EBNA) and early antigen (EA) were measured by IIF (Organon Teknika, Durham, NC, USA). IgG antibodies to HSV, VZV, and CMV were determined by ELISA (Enzygnost, Behringwerke, Marburg, Germany). All assays included negative and positive controls.

STATISTICAL ANALYSIS

Comparisons were analysed by the χ² method, or Fisher’s exact test for small numbers. The odds ratio was used to estimate the risk of being a case given the presence of exposure. In addition, a multivariate logistic regression analysis was performed with sex and age as covariables to test for any residual effects of these variables. Significance was set at p<0.05.

Results

The table shows that anti-VCA IgG antibodies were detected in 97.9% of the patients with multiple sclerosis and 81.2% of the controls (p<0.000001). IgM antibodies to VCA were present in 1.4% of the patients with multiple sclerosis compared with 0.7% of the controls (p=0.57). IgG antibodies to EA were present in 68.8% of the patients with multiple sclerosis compared with 46.5% of the controls (p=0.00007).

Anti-EA antibodies were present in 71.8% of the patients with relapsing-remitting disease compared with 55.6% of the patients with primary progressive disease (p=0.10) and 10.7% of the anti-EA positive patients with relapsing-remitting disease were in acute relapse at the time of blood sampling compared with 18.2% of the anti-EA negative patients with relapsing-remitting disease (p=0.28). Thus the anti-EA response was independent of the initial course of the disease and the disease activity at the time of blood sampling. IgG antibodies to EBNA were present in 99.3% of the patients with multiple sclerosis and 94.1% of the controls (p=0.013). Only one patient was anti-EBNA negative, but this patient tested positive for anti-VCA IgG antibodies. Thus all patients with multiple sclerosis (100%) were anti-EBNA or anti-VCA IgG positive compared with 162 (95.3%) of the controls (p=0.008).

The table shows that patients and controls did not differ in the prevalence of antibodies to other herpesviruses (HSV, VZV, and CMV).

The controls were category matched to the cases according to age and sex. However, to test any residual effect of these variables a multivariate logistic regression analysis was performed with sex and age as covariables; these analyses presented only minor changes in odds ratios and p values.

Discussion

The presence of serum IgG antibodies to EBV EBNA, or VCA, or both, indicating a history of EBV infection, were present in all the patients with multiple sclerosis and were significantly more frequent than in the control population, confirming the results of previous studies.13-17 Further, a co-twin study has reported a higher frequency of anti-EBV IgG antibodies in the twins with multiple sclerosis compared with their healthy co-twins.23 Coyleet al 19 have reported a higher frequency of anti-EBV antibodies in patients with multiple sclerosis (13 of 22) compared with controls (three of 12) although the numbers were small. The sensitivity of the test in this study was probably low because the prevalence of anti-EBV antibodies in the control group was much lower than expected.13-17 Compston et al 18 reported a lower prevalence of anti-EBV antibodies in the patient group (76%) and controls (73%). However, the patient group included patients with optic neuritis and isolated demyelinating lesions in addition to patients with clinically definite multiple sclerosis. Therefore this study is not readily comparable with the other studies with only patients with multiple sclerosis included. In addition Compston et al used other methods in their analysis (IIF) compared with the present study (ELISA) that could contribute to the different results.

The presence of anti-EA antibodies indicates acute or chronic active EBV infection and onset of viral replication.2 IgG Antibodies to EA were significantly more frequent among patients with multiple sclerosis than normal controls, compatible with the results reported by Sumaya et al.16 In the present study, two patients with multiple sclerosis and one control were anti-VCA IgM positive without clinical evidence of acute EBV infection. Further, the anti-EA response was independent of the initial course of the disease and the disease activity at the time of blood sampling. Taken together the results suggest that the patients with multiple sclerosis often have a chronic active form of the EBV infection, independent of the clinical course and activity of the disease.

No differences in antibody responses to other herpesviruses were found between the patients and controls. The prevalence of anti-HSV IgG antibodies was within the range of previous reports.14 19 24 The prevalence of anti-VZV IgG antibodies was higher than in most other studies, but confirms previous reports of no difference between the two groups.14 18 19Two previous studies of CMV IgG antibodies have shown higher frequency among the patients with multiple sclerosis than among controls, and14 19 one study found no difference between the two groups,18 compatible with the present study.

There may be several explanations for the EBV antibody pattern in the present study. It could reflect an altered immune response in multiple sclerosis or implicate EBV as an aetiological factor of the disease. An altered immune response to EBV in multiple sclerosis may reflect reduced supressor activity, which has been described in such patients.25 In addition, Craig et al 26 have reported reduced T lymphocyte mediated control of B lymphocytes infected with EBV in patients with multiple sclerosis. This may lead to frequent reactivation of the latent EBV infection, which could explain the higher prevalence of EA antibodies in the patients with multiple sclerosis. Alternatively, EBV could induce demyelination in multiple sclerosis by local infection of oligodendrocytes or other glial cells leading to release of factors causing demyelination. However, there is no evidence for such a local EBV infection in the CNS of patients with multiple sclerosis.27 Demyelination by cross reacting autoantibodies is another theory. Our recent report of anti EBNA-1 antibodies cross reacting with antigens in neuroglial cells21 is compatible with this hypothesis. Furthermore, activation of autoreactive T lymphocytes by EBV supports the hypothesis that multiple sclerosis can be initiated by a ubiquitous infection in late childhood or early adolescence that induces an immunological response in genetically predisposed people. Such an activation could occur by EBV peptides that have sequence similarity with myelin epitopes (molecular mimicry) or by EBV superantigens. To our knowledge, there are no reports of EBV peptides as superantigens in multiple sclerosis, but there are findings that can support the theory of molecular mimicry. Two studies have reported peptide sequence identities between myelin basic protein and EBV, but these were computer analyses and did not take into account secondary and tertiary structures which are of importance in the immune response.28 29 Of more interest is the report of Wucherpfennig et al 30 that EBV mimicry peptides can activate myelin basic protein specific T cell clones.

The hypothesis of cross reacting autoantibodies and molecular mimicry is attractive, but substantial evidence that such mechanisms exist in multiple sclerosis has not yet been presented. Nevertheless, with reference to the anti-EBNA antibody results in the present study, it is of interest that Bray et al 29 reported a higher frequency of anti-EBNA-1 antibodies in the CSF of patients with multiple sclerosis as well as peptide sequence identities between EBNA-1 and MBP.

A possible role of EBV infection in the complex pathogenesis of multiple sclerosis remains to be established, but the results in the present study indicate that EBV may participate in the development of the disease.