Although MS was first described in patients over 150 years ago, the etiology and pathogenesis of the disease remains under debate. The leading hypothesis in this debate, and the one that will be examined in this paper, is that MS occurs as the result of viral infection in genetically susceptible individuals. The pathology of multiple sclerosis provides the first type of evidence in support of an infectious association with multiple sclerosis. The inflammatory demyelinating lesions, abnormalities in oligoclonal bands, as well as the increased IgG levels in the CSF which are oligoclonal in type are characteristic of infectious or autoimmune disorders (Bansil). Further evidence for the multifactorial etiology and pathogenesis of MS has come from a variety of studies including: twin studies, animal and human models for virally induced demyelinating disease, and epidemiology studies. Twin studies have indicated that MS has a concordance rate of 31% in monozygotic twins and 5% in dizygotic twins (Bansil). These numbers strongly suggest that although there is a genetic component to MS, exogenous factors must also play a major role in the development of the disease. In his article, "The Pathogenesis of Multiple Sclerosis", Dr. Charles Poser states that multiple sclerosis is in fact an acquired "trait" characterized by "a permanent state of hyper-active or intensified immunocompetent responsiveness of capability" resulting from exposure to a viral antigen (either through primary infection or vaccination) in a genetically susceptible individual.

Viral causes of neurologic dysfunction have been found in several animals and also have recently been found in humans. These animal diseases, like MS, have been characterized as acute, relapsing-remitting, or progressive. One of the most widely studied animal models for MS is experimental allergic encephalomyelitis (EAE). EAE is an antigen-specific T cell-mediated autoimmune disease caused by the interaction between MHC-T cell receptor and the acetalated myelin basic protein epitope. The investigation of this disease has provided insight into the relationship between self-peptides and autoreactivity (Oksenberg). The association of viral infection and neurologic dysfunction similar to MS has recently been documented in immunocomprimised individuals. End-stage HIV patients have been shown to develop HSV associated encephalopathy. HHV-6 has also been found to associated with encephalitis and demyelinating plaques in patients with HIV disease and immunosuppressed bone marrow transplant recipients. In a 1992 study, Bray et. al stated that five individuals had neurologic disorder associated with primary EBV infection. Four of the five individuals developed acute MS within four to twelve years of their primary infection, and one developed disseminated encephalomyelitis with permanent neurological damage within two years.

EPIDEMIOLOGIC EVIDENCE

Epidemiological studies have been used to help identify possible exogenous factors which may be associated with multiple sclerosis. These studies have shown that the epidemiology of MS is very different from other autoimmune diseases such as rheumatoid arthritis and lupus, and further support the role of exogenous factors in the development of MS. The most prolific research of the epidemiology of MS has come from John F. Kurtzke. In his article, "Epidemiologic Evidence for Multiple Sclerosis as an Infection", Kurtzke describes the geographic and time distribution of MS cases throughout the world. In the United States, the pattern of MS appears to be residence based with a higher annual rate of MS associated mortality occurring in states above the 37˚ North latitude. This data also indicates that there is little difference between MS death rates in rural areas compared to urban areas; however when also categorized according to race, whites living in urban areas generally have higher rates of age-adjusted MS mortality. In Europe, the high prevalence zone of MS cases has been shown to be between 44˚ and 64˚ North latitude.

Some of Kurtzke's most extensive work on the epidemiology of MS has investigated the epidemics of multiple sclerosis on the Faroe Islands. The conclusions from these studies have helped to elucidate the nature of MS. The Faroe Islands consist of 18 major volcanic islands which are located between Iceland and Norway at 7 degrees west longitude and 62˚ North latitude. These islands were first studied in conjunction with multiple sclerosis in 1956 when John Sutherland was investigating the prevalence rates of MS in the Shetland-Orkney Island. The Shetland-Orkney Islands have a similar geography and climate to the Faroe Islands. In addition, the inhabitants of both sets of islands have an ancestry that can be traced to the invasions of the Vikings. The corresponding geographic and genetic constitution of the islands suggests that they should have similar rate of MS. However, this is not the case. In 1956, the Shetland-Orkney Islands had high rate of MS, while on the Faroe Islands cases of MS were still relatively rare.

Kurtzke continued to investigate the epidemiology of MS on the Faroe islands and discovered that there were almost no reported cases of multiple sclerosis among residents of the Faroe Islands before 1943. The validity of this observation is well supported by the characteristics of MS as well as the health care system on the Faroe Islands. Because the clinical course of MS is long term and progressive, it would be very unlikely for the disease not to be diagnosed. The prevalence of these diagnoses is increased by the fact that most unexplained neurologic disorders in young adults are first diagnosed as MS. It is very likely that all of these diagnosis are documented in records because the of the reliable documentation of medical care in the Faroe Islands. Despite this accuracy of documentation in medical records, Kurtzke took extensive measures to locate additional forms of MS documentation.

In 1974, Kurtzke began a retrospective epidemiologic study to determine the nature of the association between clinical onset of multiple sclerosis and infection by a MS-associated pathogen. The study included all MS patients who were residents the Faroe Islands and had not lived off the islands for longer than 2 years. Data was also gathered for those who had lived off the islands for three years or longer (group C) and was used in migration analysis from areas of low-risk to areas of high risk. The study participants were further divided into subgroup A, those who had never lived outside of the Faroe Islands, and subgroup B, those who had lived outside the Faroe Islands for no greater than two years. By 1986, 41 cases of MS from 102 suspected cases were identified in the Faroe Islands (25 A, 7 B, 9 C). 16 of the 32 patients from groups A and B developed MS between 1943 and 1949. The remaining 16 developed MS between 1950 and 1986. Kurtzke's later studies of MS on the Faroe Islands found that there was a fourth cluster of MS cases beginning in 1986. This data suggests that the causative agent for MS was not contracted outside of the islands, but rather is endemic to the islands themselves. While it is possible that this pathogen may have been contracted while patients left the islands for undocumented short periods of time, this is unlikely because these travels must have also occurred prior to 1940 when there were no documented cases of MS on the Faroe Islands.

Kurtzke's analyzed the epidemiology of the nine cases of MS in group C in order to determine the effect that migration from areas of low risk to areas of high risk had on the development of MS. The results of this analysis provided interesting conclusions about the incubation period of the MS-associated pathogen. They also indicated that the development of MS was associated with migration to high-risk areas after age 11. There was a time clustering of foreign residence around 10 years (with a mean of 9.3 years) before development of MS. When the two year exposure time was subtracted from this number, an incubation period of 7.2 years was determined.

Using the data from migration studies, Kurtzke was able to determine the possible time of exposure of the patients in the first epidemic and consequently identify the source of this pathogen. The first step in Kurtzke's data analysis was to categorize the subjects of groups A and B as either pre- or post-pubescent in 1943. This division caused three distinct clusters of MS to be clearly defined. The 16 patients who comprised the first cluster of MS were all post-pubescent in 1943. The second and third MS clusters were comprised of patients who were pre-pubescent in 1943. (It is interesting to note that the age of clinical onset between cluster I and clusters II and III varied significantly. Those patients in cluster I developed MS at age 30, while onset of MS occurred at age 21 in clusters II and III.) Kurtzke determined that the patients of cluster I were exposed to an MS-associated pathogen between 1941 and 1942. The main event that could account for this exposure is the British occupation of the Faroe Islands which occurred from April 1940 to September 1945. The investigation of this relationship resulted in the finding of significant temporal and geographic association between the onset of MS in all three clusters and British occupation.

The evidence which supported that the British troops introduced MS into the Faroe Islands during World War II, also supported the hypothesis that the MS-associated pathogen was a "transmissible infection". If the pathogen that the British introduced was in fact a toxin, there would not be subsequent clusters of MS. Kurtzke suggested that this infectious pathogen must usually exist in an easily transmissible but neurologically asymptomatic form which is ubiquitous in the general population. The data regarding 1943 Faroese population distribution by parishes in relation to MS onset and British occupation showed that 75.7% of the Faroese population resided in locations that housed British and/or MS patients. Kurtzke inferred from this data that 75.7% of the Faroese population was at risk for exposure to MS pathogen. The Faroese exposed to this infectious agent then later transmitted it to later generations who comprised clusters II and III.

In 1991, Kurtzke statistically analyzed the four clusters to determine if they were in fact distinct as well as the average interval between the clusters. The results of this analysis did not clearly distinguish each cluster; however, "median year of symptom onset was 1947 for epidemic I and 1986 for epidemic IV" (416). Kurtzke determined that if the distinction between the four clusters is valid, than there is an average interval of effective transmission time of thirteen years between the clusters.

In his epidemiologic investigation, Kurtzke provides convincing evidence for a viral association of MS. His migration studies also provide important information about the incubation period of this pathogen as well as the association between age of exposure to the virus and subsequent risk for developing MS. However, there are several issues, which if addressed, would not only support his study but also provide further insight into the nature of the infectious agent involved in the etiology and pathogenesis of multiple sclerosis. Criticisms of Kurtzke's results have included questions as to whether British troops introduced MS into the Faroe Islands or if, in reality there were undocumented cases of MS on the Faroe Islands. Others have commented that the incubation time that Kurtzke provides is too short, the exposure period is too long, and both are not consistent with prior epidemiological description of multiple sclerosis epidemics. The exposure period allows for the greatest amount of error in this study. If viral infection, such as EBV, is the etiological agent of MS why are two years of exposure needed. It does not take two years in a high risk area to contract EBV. Instead, it can take one five minute period exposed to the virus, or even one long kiss that may occur on a brief week-end get away. Kurtzke's study could be strengthened if he supported his results with the description of other temporal clustering of MS in other areas, such as the Shetland-Oarkney Islands as well as if these clusterings indicated similar incubation and exposure intervals. However, the greatest flaw that is apparent in Kurtzke's study is that he did not use the same sources that documented MS infection to determine probable candidates for an infection that was not endemic to the Faroe Islands prior to 1943. Although it is possible that this infection is subclinical, or that the virus introduced by the British troops was a mutated version of a virus, now associated with increased neurogenecity, which causes identical primary infection to a virus endemic to the islands, this type of analysis could at least begin to the search to identify the MS-associated virus.

Other epidemiological studies have supported Kurtzke's results. In 1991, a study by Riise et al., analyzed the space-time clustering of 381 patients with MS between 1953 and 1987 in the county of Hordaland, Norway. The results from their studies further supported Kurtzke's conclusions that a viral infection during adolescence is associated with increased risk for multiple sclerosis. Unlike in Kurtzke's study, this investigation revealed that there was a difference latency in the infectious agent. Patients were divided into birth cohorts which were defined by birth within two years of each other. The controls of this study were hospital patients who were age and sex matched to MS patients. The first step of this study was to analyze geographic clustering by determining distance between residences at age fifteen. (Age fifteen was chosen based on prior migration studies.) This residential analysis was then done for each year from birth to age 25. Each analysis excluded those patients who lived outside of Hordaland at age 15 and those who had onset of MS symptoms before age 15.

The results of this study indicated that there was a significant geographic clustering of pre-symptomatic patients between ages 13 to 20 with a marked peak of significance at age 18 (p=0.0021). The authors state that this significance could be attributed to the fact that all of the colleges and universities in Hordaland are located in Bergen, but this is not likely because this clustering was not found in control patients. The authors also state that this significant clustering before age 18 suggests that geographic significance is not due to general migration patterns in young adults. It is interesting to note that there were similar clustering patterns for patients with both remittent disease and chronic progression suggesting that there is a similar etiology for different clinical patterns of MS.

The results from this study provide even more compelling evidence in support of a virus association with development of multiple sclerosis. As stated in their report, the results from their study are comprable with infection of Epstein-Barr virus, "acquired in adolescence in genetically vulnerable persons who are also not protected by an infection acquired before this age of susceptibility. Susceptibility could be related to the route of transmission or to other age-related covariates or it may be hormonally mediated" (Riise). The epidemilogical design used to support this association was very good. However, there are problems which may have introduced error into their results. The first was that the authors were looking for peak clustering at age 15 based on prior migration studies with MS. This may have introduced bias because evaluation was not blinded. The second, which also applies to Kurtzke's study was that residence does not necessarily reveal all possible exposure. Although patients who lived outside of the county were excluded, those who traveled were still included and associations between travel locations were not evaluated. This critique is applicable to all epidemiology studies that try to evaluate exposure to an infectious agent by residence. As shown in the hantavirus pulmonary syndrome epidemic in the four corners region during 1993, exposure can occur during one hike. Although the mode of transmission of EBV and hantavirus are different, the fact still remains residence does not adequately evaluate exposure to a virus. For this reason, future studies should try to determine vacation locations, ex. summer camps or resorts, and determine if EBV or another virus occurred at time of this travel. Scientists could also confirm the presence of infection through antibody analysis.

Many scientists have used antibody analysis to try to identify possible MS associated infections. Recently these antibody studies have investigated the association between multiple sclerosis and Epstein-Barr virus (EBV) and human herpes virus-6 (HHV-6). EBV is a human herpes virus known to be associated with subclinical infection in young children an d the development of mononucleosis in adolescents and adults. Human herpesvirus-6 (HHV-6) is the causative agent in roseola exanthum subitum, a common febrile illness in children. Recently HHV-6 has also been found to be associated with encephalitis and demyelinating plaques in patients with HIV disease and immunosuppressed bone marrow transplant recipients. The infection of HHV-6 is widespread with a seroprevalence rate among healthy adult populations between 50% and 85%. This seroconversion has been demonstrated to occur early in childhood, with antibody titers decreasing in the population by age 40. Both viruses are good candidates for association with MS because characteristics of infection may be consistent with epidemiological description and both have been found to be associated with neurologic disorder.

Problems with forming conclusions from antibody investigations are exacerbated by the fact that many of the viruses being tested are very ubiquitous in the population, with seroprevalence rates reaching 89% in the case of EBV. However, as discussed in a report by Bray et. al, despite the high prevalence in the general population, there a significant difference between the seroprevalence of EBV in MS patients compared to controls (100% vs. 89%; p<0.0001). In addition, the Ab titers against viral capsid proteins was significantly higher in MS patients (p<0.0001). However, it is important to recognize that the increased antibody titers in patients with multiple sclerosis compared to controls may be evidence that the immune system becomes compromised in multiple sclerosis, not that that particular virus is involved in the etiology and/or pathogenesis of the disease.

In the article, "Human herpesvirus 6 and multiple sclerosis: survey of anti-HHV-6 antibodies by immunofluorescence analysis and of viral sequences by polymerase chain reaction," evidence in support of HHV-6 as the MS-associated pathogen is given. The authors also analyzed the relationship between Ab titers and presence of viral sequence in order to determine whether the cause of increased Ab titers was the result of reactivation of latent infection rather than immune impairment. Serum was collected from 126 patients with multiple sclerosis, (47 men, 79 women, mean age 35 years, range 13-70) and 500 HIV-seronegative blood donors (250 men, 250 women, mean age 44 years). These serum samples were tested for anti-HHV-6 antibody by indirect immunofluorescence analysis (IFA). PCR was also performed on the DNA extracted from the peripheral blood mononuclear cells (PBMC) of 31 MS patients and 24 blood donors.

The results of this study indicate that higher anti-HHV-6 IgG antibody titer was found in the serum of MS patients compared to the serum of control subjects (p<0.005). Although the presence of HHV-6 Ab in CSF was not determined for blood donors, 7% of MS patients showed presence of the Ab in undiluted samples. Despite this high antibody level, PCR of the PBMC DNA testing for specific primers in HHV-6 only revealed positive results in 1/31 MS patients and 1/24 normal subjects.

The authors attributed the high levels of anti-HHV-6 antibody titers in the serum of MS patients to cellular immune system impairment leading to reactivation of HHV-6 latent infection. But as shown by PCR data, this reactivation is in the absence of detectable viral sequences. For this reason, the authors also state that the presence of anti-HHV-6 antibody in the CSF of MS patients is caused by immune stimulation leading to the synthesis of antibodies by intrathecal lymphocyte clones.

There are many problems in this experiment which are applicable to this and many other studies investigating MS-associated pathogens by the presence and levels of antibodies. Future studies should include "refined antibody test systems, experience with larger numbers of subjects, and careful prospective and molecular studies" in order for conclusions about the association between HHV-6 and MS to be determined (Sola). These studies should also focus on the presence of HHV-6 in CSF of patients and only use antibody data from serum samples to confirm results. It would be very interesting if in the study of CSF samples of MS patients, patients with HHV-6 encephalopathy were included with healthy subjects as controls.

DELAYED CHILDHOOD INFECTIONS

While trying to identify the MS-associated pathogen, scientists have also found evidence in support of an increased risk for onset of multiple sclerosis with adolescent viral infection. In "Infections in Childhood and Adolescence in Multiple Sclerosis", Gronning et al. investigated the association between delayed childhood infections and the development of MS. The study consisted of one hundred fifty-five people with MS and 200 controls. The age-, sex-, and residence-, controls were patients with diagnoses ranging from traumatic rupture of ligaments to plastic surgery disorders. All subjects filled out a questionnaire relating to childhood infectious diseases (measles, rubella, chicken pox, mumps, whooping cough, and scarlet fever) as well as respiratory infectious diseases. The subjects were then interviewed by one of the authors. The associations between MS and different viral or bacteriological infections during childhood were then analyzed in a logistic regression model.

The results of this study indicated that the proportion of reported cases of measles, rubella, chicken pox, and mumps was similar between MS patients and controls. Analysis did show that there was a higher proportion of whooping cough and scarlet fever in MS patients compared to controls. There also appeared to be a slight statistical significance between the mean age of measles in MS patients compared to controls (6.7 years vs. 5.7 years, P=0.056). There was a linear trend in this association that was close to being statistically significant, suggesting that developing measles in adolescence is associated with increased risk for MS. No statistically significant association was shown between common bacterial infections including, otitis, tonsillitis, sinusitis, bronchitis and pneumonia. However, there was a significant increase in the number of tonsillectomies performed on MS patients than in controls.

The authors of this article claimed that this study provides convincing evidence in support of an increased risk of MS in groups where childhood infections were age delayed. However, this was not supported by their data. The only statistically significant difference between the MS patients and control groups was the mean age of development of measles infection. In addition, the difference in mean age of development was only one year and both were under the age of seven, which does not support the hypothesis that development of childhood infection during adolescence increases the risk for MS. It was interesting to note that the there was also a statistically significant difference in the number of tonsillectomies between MS patients and controls. This difference was especially interesting because of the effect of tonsillectomies on the immune system. Tonsillectomies have been shown to cause an decrease in serum immunoglobulins and secretory IgA as well as an increase in allergic rhinitis in children for a period of 1-4 months after surgery. This change in the immune system may cause a child to be increasingly susceptible to an MS-associated pathogen by changing the pathogenesis of childhood infection.

There are several problems in this study with weaken the conclusions that the authors formed from their data analysis. The major flaw of this study was that the data was reliant on subject self-reporting. The subjects were required to recall the exact age that they developed typical childhood infections. It would be very hard for any person to recall specific years, let alone individuals who may have some degree of neurocognitive dysfunction. Therefore it is possible that the statistical significance found between age of measles development in MS patients and control group may in fact be the result of problems in memory recall. It was also interesting that EBV was not included as a common infection. When this paper was written, scientists were beginning to propose that EBV was the MS-associated pathogen. This hypothesis is supported by epidemiological studies which suggest that the MS-associated pathogen causes infection between 16-20 years of age. EBV has been shown to have very different manifestations in different populations. In lower SES groups, EBV infection occurs early in childhood and is often sub-clinical. However, in higher SES groups EBV occurs after puberty and often manifests as mononucleosis. Future studies must address this issue. Ab tests, in addition to self-reported data should be used. If the individual does not recall EBV infection, it may be assumed that subclinical infection occurred possibly during early childhood. However, if infectious mononucleosis is reported, age of occurrence must also be noted. This data should then be analyzed and increased risk for MS in individuals with EBV infections could be determined.

In their pursuit to prove the association of viral infection in the etiology and pathogenesis of multiple sclerosis, scientists have formed three main hypotheses explaining the mechanism of this involvement. The first of these hypothesis, and the most widely discussed, is that molecular mimicry between viral antigen and myelin basic protein activates the autoimmune mechanism of multiple sclerosis. The hypothesis that the autoimmune mechanism of MS is activated through molecular mimicry has been supported by studies which have found homologous proteins in antigenic regions of Epstein-Barr virus and myelin basic protein. The second hypothesis is that infection of a virus during adolescence could establish a latent infection in the central nervous system. Scientists propose that reactivation of this latent virus could lead to "cytopathic and/or immunological damage to oligodendrocytes" (Challoner et al.). A third hypothesis is that viral infection may cause exacerbations of multiple sclerosis leading to the further development of lesions. This influence on exacerbation rate may be through the activation of immune host response leading to up-regulation of the immune system or alteration of the permeability of the blood-brain barrier.

Epstein-Barr virus is a human herpes virus. It is has been shown to replicate in the epithelial cells of the nasopharynx and salivary glands. In latency, EBV usually exists as a plasmid in B lymphocytes. EBV has been shown to cause infection in 80% of the population. It has also shown to have different clinical manifestations based on age of infection. In young children EBV infection is often subclinical; however, in children 15 years or older, EBV infection manifests as mononucleosis. EBV infection has also been associated with Burkitt's lymphoma, nasopharyngeal carcinoma, and several disorders involving neural complications. These characteristics of EBV have caused it to be implicated as the possible MS-associated infectious agent.

The age-dependence of EBV infection on clinical disease, agree with the viral model proposed by epidemiological evidence. In his epidemiological report, Kurtzke suggests that primary infection of MS pathogen must be fairly ubiquitous within the population. He also states that viral infection in adolescents is associated with increased risk for onset of MS. These two descriptions fit within the model of EBV which has been shown to cause infection in 80% of the general population. The association between EBV and MS was further supported by the occurrence of five individuals who had neurologic disorder associated with primary EBV infection. Four of the five individuals developed acute MS within four to twelve years of their primary infection, and one developed disseminated encephalomyelitis with permanent neurological damage within two years (Bray).

In 1992, Bray et. al investigated the association of EBV with MS. The purpose of this study was to analyze the antibodies to EBV using three different assays in patients with MS as well as identify pentapeptide similarities between EBNA and myelin basic protein. Study participants included patients from a pool of 245 individuals whose diagnosis and laboratory findings were available. MS patients were diagnosed according to standard clinical criteria and were only included in the study if the BBB was shown to be intact. Control subjects were patients with clinical diagnoses ranging from progressive nondemyelinating degenerative disorders to psychiatric disorders, and were only included if they were seropositive to EBV, had an intact BBB, and had no presence of demyelination. Quantitative and qualitative analyses of CSF IgG were determined using Western immunoblots and presence of oligoclonal bands in CSF was analyzed using agarose gel electrophoresis. Several cell lines were used as sources for EBV antigens including viral capsid antigen which was later identified using anticomplement immunofluorescent assays. Enzyme-linked immunosorbent assays (EIAs) were also performed using purified EBNA-IgG protein to test for target anti-EBNA antibodies in the CNS. Analyses of the 70kD protein occurred using cell extracts from cell lines infected with EBV. Possible sequence homologies between EBNA-1 and myelin basic protein were then searched for using FASTA program.

The results of this study showed that 35/50 (70%) of MS patients and 8/50 (16%) of NDC had anti-EBNA Abs present in the CNS (p<0.001). When EIA was used to test specifically for the presence of anti-VCA Abs, antibodies were detected in 24/50 (45%) MS patients and in 14/50 (28%) of controls (p<0.05). Assays testing for the presence of 70 kD protein in CSF samples showed that 79/93 (85%) of MS patients were positive for the presence of the protein. However, MS patients who did not show presence of this protein had oligoclonal IgG bands in the CSF. This protein was only shown to be present in 13% of NDC samples using ACIF. Antibodies to VCA were found to be present in 32/47 (68%) of MS CSF samples and 12/5 (22%) NDC. Further analysis indicated that geometric titers of anti-VCA in CSF for 77 MS patients was 2.79 compared to 1.41 in 41 NDC (p<0.001). The FASTA program which searched for homologous peptide sequences between EBNA-1 and 32,000 other protein sequences resulted in two identical matches. These identical matches were 5-amino acid sequence in EBNA-1 (641 AAs) and myelin basic protein (169 AAs) referred to as QKRPS and PRHRD.

The results of this study indicate that the CSF of patients with MS contains excessive amounts of anti-EBNA-1 antibodies compared to controls. Evidence in support of this included detection of high anti-EBNA-1 Ab titers using EIA with EBNA-1 purified protein as a target antigen in addition to reactions of CSF with 70 kD protein on Western Blots of Raji cells know to be infected with EBV. These results have future implications in terms of laboratory tools which could be used to diagnose MS and possible mechanisms of the pathogenesis of MS. However, as shown in the discrepancy between the percent of MS patients with anti-VCA Ab using EIA compared to ACIF, the assay used has a direct effect on diagnosis.

In a more recent study published in the Journal of Immunology, Vaughan et. al used Enzyme-linked antibody assays to illustrate that infection with EBV produces antibodies that cross reacted with 80-82 kDa and 60 kDa proteins in neuroglial cells. These protein sequences may consist of the pentapetide sequences that Bray et al. concludes are also found in EBNA-1. As in Bray's study, the results from Vaughan's study indicated that MS patients express increased antibody titers to EBNA-1 in their sera during exacerbations. These antibodies were shown to be specific to the glycine/alanine repeat in EBNA-1.

The results of these two studies support the hypothesis that EBV may cause activation of the autoimmune mechanism through the process of molecular mimicry. They also implicate that the immune system is involved in the pathogenesis of multiple sclerosis. The presence of elevated anti-VCA Abs in the CSF may be a reflection of a small population of EBV-transformed B cells entering the lytic cycle. In addition, the homologous sequences between pentapeptides in myelin basic protein and EBNA-1 suggest that T-lymphocytes play an specific role in the demyelination of neurons.

The current hypothesis for the autoimmune mechanism of multiple sclerosis is that the dymelination of plaques is mediated by macrophages and T lymphocytes. While the molecular mimicry hypothesis explains that these cells may have a sensitization to myelin basic protein because of homologous sequences found on antigenic viral proteins, it is also possible that this sensitization may be the result of persistent viral infection within the CNS that when activated, leads to cytopathic and/or immunological damage to oligodendrocytes. A resent study by Challoner et al. published in Proceedings of the National Academy of Science, suggests that HHV-6 may be involved in the etiology and pathogenesis of MS via this mechanism. Human herpesvirus-6 (HHV-6) is the causative agent in roseola exanthum subitum, a common febrile illness in children. Prior studies have indicated that after roseola infection, HHV-6 is often detected in the central nervous system. In addition to its association with encephalopathy and neurotropism, HHV-6 is a interesting candidate for a multiple sclerosis-associated pathogen because it has a strong CD4+ T cell tropism and its infection profile is consistent with the epidemiologic evidence for multiple sclerosis.

In Challoner's investigation, representational difference analysis (RDA) was used to detect HHV-6 in the brains samples from MS patients and control patients who died of either neurological disorders or trauma cases. RDA analysis involves "successive rounds of subtractive hybridization and PCR amplification enriched for DNA sequences that were present in a DNA preparation from diseased tissue (MS patients) but absent from control DNA (nondiseased cases)" (Challoner). In addition to RDA, nested PCR was used to detect HHV-6 in samples. Samples also underwent immunocytochemistry assays in order to detect the presence and location of HHV-6 as well as other herpesvirus thought to be probable candidates for the MS-associated pathogen.

The results of the RDA analysis indicated a 341-nucleotide fragment which was 99.4% identical to HHV-6 variant B major DNA binding protein (MDBP). HHV-6 was found in 25/32 MS specimens and 40/54 controls. DNA sequencing indicated that 36/37 (both MS and control samples) were positive for HHV-6 variant B group 2. Quantitative PCR analysis showed that 47/49 MS samples had HHV-6 viral titers comparable to controls (2 had significantly higher titers). Immunocytochemical localization assays indicated identical patterns for two HHV-6 proteins in MS samples. HHV-6 nuclear staining of oligodendrocytes was shown in 12/15 MS samples but not in any control cases (0/45; P<0.0001). This nuclear staining was more common in plaque regions compared to normal white matter (27/28 vs. 10/18; P<0.0001). Plaques were shown to be heterogeneous for nuclear oligodendrocyte staining. The gray matter adjacent to plaques also showed prominent cytoplasmic staining. Nested PCR used to screen for other herpesviruses HSV-1 in 2/25 MS cases and 2/42 controls and VZV in 0/17 MS cases and 2.24 controls. These results suggest that HHV-6, and not EBV, is the most probable human herpes virus to be associated with the etiology and pathogenesis of MS.

This was a well designed study in terms of the number of different assays used to test for possible MS-associated pathogens. I found it very compelling that the authors went beyond using PCR to test for the presence of HHV-6 in the brain. By staining the ogliodendrocytes and determining the location of HHV-6, the researchers were able to indicate that HHV-6 is directly involved in the formation of plaques. The immunochemistry assays were also necessary because if HHV-6 is so ubiquitous and has latent infection in the brain, it would be expected that the majority of samples (controls and MS patients) would be positive for HHV-6. This study could be strengthened if all samples were included in each assay. The authors had access to 246 samples from MS patients, but only 50 were used in each assay. The analysis of all samples may help to further support the differences that were shown. Although the differences between the staining of oligodendrocytes in MS vs. control cases and plaque regions vs. normal white matter were found to be statistically significant.

INFLUENCE OF INFECTION ON EXACERBATIONS OF MULTIPLE SCLEROSIS

In addition to being directly involved in the etiology of MS, viral infections may also cause exacerbations, suggesting that viruses are also indirectly involved in the pathogenesis of the disease. Studies have shown that the exacerbations of the lesions of multiple sclerosis can be attributable to exogenous events such as stress, trauma, infection, immunization, pregnancy, climatic changes, and physical exertion. One of the best documented exogenous events leading to exacerbation is upper respiratory viral infection (URI). Results from studies investigating the relationship between URIs and exacerbations have suggested URIs may cause an increase in the production of interferon-gamma (IFN-gamma) which in turn may up-regulate the immune system and cause it to enhance the autoreactive response to the CNS. It is also possible that the increase in production of IFN-gamma may alter the permeability of the blood brain barrier (BBB) allowing the immune system to have greater access into the CNS. If it is true that there is an association between exacerbations of multiple sclerosis and increased IFN levels, than it is likely that exacerbations are not associated with one particular virus, but rather general viral infection.

In 1994, Hillel S. Panitich published the results of his investigation which assessed the association between URI events and multiple sclerotic exacerbations. Thirty patients were included in this prospective study. These patients were participants in a concurrent study evaluating the effectiveness of different doses of IFN-beta in reducing events of multiple sclerotic attacks. Each patient kept a daily log in which they recorded the symptoms of infection in themselves, family members, friends, and co-workers. These subjects were evaluated every three months, or upon attack of MS, and tested for the presence of antibodies common to upper respiratory infections. Subjects were placed into "at risk" and "not at risk" groups. Periods of "at risk" were determined as being 1 week prior and 5 weeks after documented infection. Therefore any MS attack that took place during that time was proposed to be associated with infection.

The results indicated that there was a strong correlation between MS attacks and URIs. Two-thirds of exacerbations occurred during periods "at risk" and only one third of exacerbations occurred during periods "not at risk". The rate of attacks was shown to be independent of treatment with IFN-beta but varied between "at risk" (2.92 attacks/year) and "not at risk" groups (1.16 attacks/year). However these attacks were not qualitatively different. There were no differences in seasonal patterns between the risk groups. There were also no differences in seasonal patterns of infection between treatment groups; however, the pattern of exacerbations was different than that for infection suggesting that IFN treatment does not have any effect on the occurrence of infections but does prevent the occurrence of exacerbations. There was a marked decrease in the number of MS attacks in patients receiving the highest dose of IFN-beta. Tests on serum antibodies showed that they were from various pathogens, but were predominantly viral. It is interesting to note that EBV Abs fluctuated parallel to attacks but the authors stated that this was a result of non-specific activation.

The authors stated that this association between viral infection and MS exacerbations is most likely due to the viral infection causing the immune system begin a systemic release of IFN-gamma and subsequent release of TNF which further up-regulates the immune system and allows T cells to enter the CNS. This entry would then cause an increase in multiple sclerotic exacerbations. The IFN-beta treatment tested for efficacy in this study is thought to decrease the immune response to viral infection and therefore cause a subsequent decrease in the amount of MS exacerbation. The data supported the authors' hypothesis in that IFN-beta treatment was shown to decrease the amount of MS exacerbations but not the frequency of viral infection.

There are several significant problems with this study which weaken the validity of its results. The first of these problems was the small number of subjects included. This problem was exacerbated by the fact that the purpose of the study was not only to asses the correlation between viral infection and exacerbations of MS but also to determine the efficacy of IFN-beta treatment. Statistical significance was reported regarding the association between number of exacerbations during periods at risk and periods where the individual was not at risk, however no other statistical analysis was provided. Another problem with the study was that although periods at risk were identified, titers of corona virus and rhinovirus were never performed and actual exposure to virus was never proven. Viral serology was performed on the subjects. The authors claim that the high variation between viral antibodies shown indicates a non-specific viral infection; however it is also possible that the results correlate to previous infection not current exposure to viral pathogen. Overall, this was a very interesting study, but its over-ambitious nature prevented any conclusive evidence to be determined.

A recent study published in the Journal of Neurology, Lyke et al., investigated the association between viral infection and MS using the antiviral therapy acycloguanosine. Acycloguanosine, commonly known as acyclovir, is a nucleoside analog which inhibits viral DNA polymerase, thereby blocking the synthesis of viral DNA. The purpose of the study was to determine whether acyclovir-sensitive viruses were involved in the development of multiple sclerotic exacerbations. Sixty patients with MS were included in this study. These patients had been followed prospectively for two years prior to the beginning of the trial. MS patients were split into two groups bases on exacerbation rates during the two-year period before commencement of the trial. These groups were then randomized to receive either placebo or acyclovir treatment. The acyclovir treatment consisted of 800 mg given 3 times daily for 2 years. The exacerbation rate served as the primary response variable. Exacerbations were also classified according to severity and later according to frequency. In addition to these evaluations, neurological examinations, neuro-opthalmological examinations, neurophysiological examinations, and neuropsychological examinations were performed. Ab tests were also performed on serum and CSF samples from all patients. These tests detected for the presence of Abs specific for HSV-, HSV-2, VZV, CMV, and EBV. Antibodies to HHV-6 were not evaluated.

The treatment group was shown to have an exacerbation rate of 1.03. The exacerbation rate of the placebo group was determined to be 1.57. Because this data was statistically insignificant, the authors reanalyzed the exacerbation according to frequency in each treatment group and statistical significance was shown (P=0.017). Exacerbation rates in HSV seropositive patients in treatment group was 0.97 compared to rate in placebo group of 1.31. The effect of acyclovir was also analyzed based on differences in exacerbation rates between the two year pre-study period and treatment period and a significant reduction in frequency was determined (P=0.024). Acyclovir was shown to have a reduced mean annual exacerbation rate of .44 while the placebo group had an increased mean annual exacerbation rate of .27. Neurological examination showed increasing deficit in both groups and acyclovir treatment was not shown to inhibit the spread to MS lesions. Acyclovir treatments had significant reduction (P=0.002) in antibodies to HSV- and HSV-2 over treatment period. The number of antibodies was also shown to be significantly lower than in the placebo group (P=0.046). However, acyclovir was not shown to affect antibody levels in the CSF. Acyclovir was also shown not to affect VZV, EBV, or CMV antibody titers.

The authors provide convincing evidence for the decrease in exacerbations in MS patients receiving acyclovir. Although, this reduction was not found to be statistically significant until the exacerbations were classified according to frequency, the treatment still appeared to be effective. It was interesting to note that there was an increase in the number of exacerbations in the placebo group during the treatment period compared to the prior 2 year study period. It was also interesting to note that this increase was almost equal in value to the decrease found in the acyclovir recipients. This data seems to suggest that the placebo group may have actually caused an increase in the amount of exacerbations. Although the evidence supporting the effectiveness of acyclovir was very strong, it is important to acknowledge that the treatment was specific to HSV- and HSV-2. This specificity for HSV may bring in questions of blindedness. Because acyclovir is effective in treating oral and genital herpes blisters, it is possible that patients with these conditions would have been able to determine which treatment they were receiving. However, when asked to guess which regiment they were receiving, there was not a significant difference in their ability to guess. The authors state that because acyclovir did not have an effect on the progression of neurological deficit, acyclovir-sensitive viruses are not a cytotoxic agent in MS. It is very likely mechanism for involvement is that the acyclovir treatment prevented primary infection of a virus or the reactivation of acyclovir-senstive virus. Both events could possibly lead to a change in the permeability of the blood brain barrier which would have allowed the immune system's penetration of the CNS and consequent attack on the protective Schwann cells which contain myelin basic protein. It must also be acknowledged that the affect of acyclovir may not be through viral manipulation but instead direct influences on host responses. However, due to specificity of acyclovir on human herpes virus, this association is unlikely.

The investigations discussed in this essay support the hypothesis that viral infection is involved in the etiology and pathogenesis of multiple sclerosis; however convincing proof of this association does not exist. One compelling study can never provide unquestionable reliability of this association and therefore investigational methods must be repeated in order to increase the validity of results. To date, the most prominently repeated studies have involved the epidemiology of MS and antibody analysis. The progress of finding the association of viral infection and MS is dependent on communication within the scientific community. Unfortunately, there is no evidence that this exists. For example, the two articles which suggest that homologous sequences between antigenic EBV protein and myelin basic protein were separated by a period of four years. No other studies on this subject were published during this time. The research on the association between MS and HHV-6 is also very disappointing. The research by Challoner provides compelling evidence that infection by HHV-6 may lead to cytopathic or immunological effects in the CNS; however, the most recent article of HHV-6 by Gaer and Driessche to be published in the next issue of the American Journal of Neurology merely reports the finding that patients with MS have increased levels to HHV-6 in the CSF.

Because of the strong evidence in support of the association of EBV and HHV-6 in the etiology and pathology of MS, future studies should further investigate whether each virus may be associated with different clinical manifestations of multiple sclerosis. Although Riise's epidemiology study suggests that different patterns of MS share common etiologies it is also very likely that MS is a syndrome with several distinct etiological agents. In order to study this relationship, MS patients with acute, relapsing-remitting, and progressive in addition to healthy controls should all be included in future studies.

A convincing study showing the association between HHV-6 or EBV with multiple sclerosis should address role of viral infection on the etiology as well as pathogenesis of MS. The first step of the study should be to investigate the relationship between age of primary infection and development of multiple sclerosis. This would be especially useful in determining association between MS and EBV. Antibody studies in conjunction with self-report data could be used in this analysis. If subjects state that they never had mononucleosis but have Abs to EBV, it may be that they had a subclinical primary infection most likely would have occurring before puberty. According to the current hypothesis, this type of primary infection should be more prevalent in controls compared to MS patients.

The second step of the study would be to test for the presence of viral DNA in the cells of the CNS. The study design of this step of the study could closely resemble that used in Challoner's experiment which used RDA, nested PCR, and immunocytochemistry assays.

The third step of the study would test the mechanism for the association between EBV and MS. It should model the methods used in Bray's research which determined that there were homologous peptide sequences in EBNA-1 and myelin basic protein. In addition, this step should also use the methods of Challoner to determine if in fact EBV has latent infection in the CNS in patients with MS.

Once more evidence is compiled in support of a viral role in the etiology of MS and the possible mechanism for this involvement, investigations should also begin to analyze the indirect role of viral infection in the exacerbations of multiple sclerosis. The results from these investigations will not only help scientists to better understand the intricate relationship between viral infection and autoimmune disease, but more importantly the results from these investigations may help the development of new therapies that could be used to treat those living with multiple sclerosis.