DISEASE HETEROGENEITY

The heterogeneous nature of multiple sclerosis is reflected by its variable clinical phenotype, its nonuniform neuropathology, and its heterogeneous molecular pathogenesis. Both genetic and environmental factors are believed to have an effect either in modulating susceptibility or influencing the development of the disease. Autoreactive T cells are considered to play a key role in mediating the disease process. Evidence for a role of autoreactive T cells stems from the composition of inflammatory infiltrates in the CNS, which consist mainly of lymphocytes and monocytes, and from data from its animal model, experimental allergic (autoimmune) encephalomyelitis. The injection of myelin components into susceptible animals leads to a CD4⁺ T cell–mediated autoimmune disease resembling multiple sclerosis and can be adoptively transferred from sick to healthy animals via encephalitogenic CD4⁺ T cells. A role of autoaggressive T cells in multiple sclerosis is further supported by the therapeutic, although limited, efficacy of immunosuppressive and immunomodulatory agents and by the fact that certain major histocompatibility complex (MHC) class II alleles represent the strongest genetic risk factor, presumably because of their role as antigen-presenting molecules to pathogenic CD4⁺ T cells.

Inflammatory events are considered to initiate and drive the disease process during early stages. The myelin damage and axonal injury that account for the permanent neurological deficit seen during later phases of multiple sclerosis probably result from a complex sequence of events, including processes intrinsic to the CNS, such as increased vulnerability to tissue injury and/or poor repair, which might progress independently of immune factors. Multiple sclerosis is therefore not solely a disease of the immune system; CNS-specific components, although largely overlooked in their potential pathogenetic role, are presumably equally important for its pathogenesis.³

The clinical pattern of multiple sclerosis is generally divided into two major forms. The first, most frequent (85% to 90%) subtype follows a relapsing-remitting course and is characterized by separate episodes of neurological deficits involving different sites of the CNS, each lasting for at least 24 hours and separated by intervals of at least 1 month. Relapsing-remitting multiple sclerosis usually evolves over decades and in most cases transforms into a secondary progressive course. About 10% to 15% of patients present with insidious disease onset and steady progression, termed primary progressive multiple sclerosis. There is heterogeneity in morphological alterations of the brain, as visualized by either magnetic resonance imaging (MRI) or histopathological evaluation, but also in clinical presentation, such as which CNS system and areas are primarily affected and whether a patient responds to treatment. The factors underlying the different disease courses and the disease heterogeneity are incompletely understood but presumably include a complex genetic trait that translates into different immune abnormalities and/or increased vulnerability of CNS tissue to inflammatory insult or reduced ability to repair damage.

Current knowledge about how certain genes confer risk for multiple sclerosis or any other autoimmune disease at the molecular level is incomplete. However, numerous studies on the genetic epidemiology of multiple sclerosis, which are described in the following sections, provide compelling evidence that the susceptibility to the disease is inherited, although additional environmental triggers might be necessary to translate disease susceptibility into the clinical phenotype.

POPULATION PREVALENCE

The disease prevalence varies between 60 and 200 per 100,000 in North America and northern Europe and generally follows a north-to-south decreasing gradient on the northern hemisphere and the opposite on the southern hemisphere, with very low rates or virtual absence of the disease near the equator (Fig. 74-1).

Figure 74-1 Worldwide distribution of multiple sclerosis. Multiple sclerosis is most common in parts of the world with cooler climates. Europe, North America, and parts of Australia and New Zealand have multiple sclerosis prevalence rates (60 to 200 per 100,000) that are much higher than rates in countries clustered around the equator.

(Adapted from Marrie RA: Environmental risk factors in multiple sclerosis aetiology. Lancet Neurol 2004; 12:709-718.)

This geographical distribution can be attributed to both environmental factors and genetic effects. For many years, an infectious etiology of multiple sclerosis has been suspected, because it is consistent with a number of epidemiological observations and with immunopathological characteristics of the disease. Migration studies showed that individuals who migrate from high-risk to low-risk areas after the age of 15 tend to retain their risk of multiple sclerosis, whereas individuals who migrate from high-risk to low-risk areas before the age of 15 acquire a lower risk; this indicates that childhood exposure to an environmental factor increases disease susceptibility. Supportive data for an infectious agent also come from reports of endemic clusters of multiple sclerosis. After the British occupation of the Faroe Islands, off the coast of Denmark, where no cases of multiple sclerosis had been reported before, several islanders developed the disease between 1940 and the end of World War II, and the affected areas were found to be locations of British troop encampments after 1940. Other examples of multiple sclerosis epidemics have been described in northern America and Europe. These observations suggest that an environmental factor is relevant for the initiation of the disease process, and in the established disease, infections are additionally known to be capable of triggering exacerbations.

No specific transmissible agent has so far been linked convincingly to multiple sclerosis. The most consistent evidence of a potential role in the disease exists for Epstein-Barr virus and human herpesvirus 6, as a result of the detection of viral DNA in brain specimens derived from multiple sclerosis lesions (in the case of human herpesvirus 6) and by convincing seroepidemiological studies. Both are ubiquitous viruses that act at the population level and produce latent, recurrent infections. Generally conceptualized as a trigger for the manifestation of the disease in genetically susceptible individuals, the mechanisms by which these viruses and other potential candidates initiate, exacerbate, and perpetuate the disease are, however, far from understood.

Arguing for a genetic basis of multiple sclerosis is the fact that the disease prevalence differs strikingly between geographically close but genetically distinct populations. Ethnic groups such as the Lapps in Scandinavia, Gypsies in Hungary, Maoris in New Zealand, or Aborigines in Australia are rarely if ever affected by multiple sclerosis, although the disease is otherwise common in these latitudes. Furthermore, multiple sclerosis is rare among Japanese and Chinese populations, African Blacks, North and South Amerindians, and the native populations in southern countries of the former Soviet Union (Turkmen, Uzbeks, Kazakhs, Kyrgyzis), but it occurs notably more frequently among Whites living in the same area. Further examples are the different prevalence rates in genetically distant populations living on the same island, as reported for Sardinia, Cyprus, and Ireland.

The idea that genetic factors may have a role in multiple sclerosis was first raised in the 1890s with the identification of familial aggregation. Further milestones in establishing the concept of genetic susceptibility to multiple sclerosis were the appreciation of ethnic risks in the 1920s, comprehensive studies on the geographical distribution of the disease in the 1950s, and the description of the MHC associations in the 1970s. It took, however, a century until the first systematic genetic epidemiological analyses on familial aggregation were initiated. These studies formed the groundwork for the understanding of multiple sclerosis as a genetically determined disease.

FAMILIAL AGGREGATION

Familial aggregation studies, including research with twins, siblings, and adoptees, demonstrated that the risk of developing multiple sclerosis increases with the degree of relatedness between individuals.⁴ For example, monozygotic twins of patients with multiple sclerosis have a risk more than 100 times higher than that of the general population of developing the disease, and full siblings have a lifetime risk approximately 20 times higher than that of the general population (Table 74-1). Although these recurrence risk values are considerably lower than the ones for mendelian-dominant disorders such as Huntington’s disease (approximately 5000-fold increased risk for siblings), they are similar to the risks that have been reported for other complex polygenic diseases, such as systemic lupus erythematosus or type I diabetes mellitus (approximately 20- to 30-fold increased risk for siblings).

SUSCEPTIBILITY GENES

Numerous searches have been performed and are currently ongoing in order to identify single candidate genes or a set of genes that are relevant for the disease. The strongest and most consistent evidence exists for the area of the MHC, the human leukocyte antigen (HLA), on chromosome 6p21. The total genetic susceptibility attributed to the HLA locus in multiple sclerosis is estimated to be between 15% and 50%.⁷

Other candidate genes selected for their potential involvement in the pathogenesis of multiple sclerosis—such as T cell receptor genes, genes encoding myelin components, interferons, various chemokines, cytokines, and complement factors, to name a few—have been studied extensively. The results of these studies, however, were inconclusive and most of the initial positive reports could not be confirmed in subsequent analyses.

In an alternative approach to a traditional case-control design that has been applied to investigate allelic variants of single candidate genes, whole-genome screens of affected relatives have been used to assess linkage of polymorphic markers spread throughout the genome. Linkage studies have been successfully applied in genetic diseases that follow a mendelian single-gene trait with a high penetrance of the allelic variant and a clear clinical phenotype, such as those in Huntington’s disease or in muscular dystrophy.

Linkage studies in multiple sclerosis, mainly performed in North America and northern Europe, did not provide strong evidence for a single major contributing locus, but, in accordance with the previously mentioned surveys on single candidate genes, they consistently identified an increased sharing of the MHC region (6p21) in affected individuals. No consensus could be reached for other susceptibility loci.

The ambiguity of gene searches in a complex and heterogenous disease such as multiple sclerosis is presumably at least partially a consequence of methodological problems such as an insufficient stratification of the population investigated with regard to their HLA haplotypes, clinical-pathological features, or ethnic background. In addition, the classic genetic epidemiological methods that have been applied so far are probably of limited power to identify and localize small genetic effects for complex traits with unknown modes of inheritance. Furthermore, even though whole-genome screens employed 6000 microsatellites, these are still too far apart and unevenly spread out through the genome.

Because no major single risk locus could be identified by whole-genome linkage studies, a large number of common allelic variants, each with only subtle but important variations, might synergistically lead to the major genetic risk that is associated with the occurrence of the disease. For example, the overall number of susceptibility loci that have been identified in the most comprehensive genomic screen so far is 411.⁸ Because of the low resolution and the weak signals that can be achieved by traditional linkage studies in polygenic diseases, the identification of such common variants requires different approaches. To date, the MHC class II region remains the only area of the whole genome clearly associated with multiple sclerosis, although the precise genes within this area that confer disease susceptibility are not yet known.

THE HUMAN LEUKOCYTE ANTIGEN GENE COMPLEX

The MHC region on chromosome 6p21 encompasses 4.5 megabases and encodes approximately 200 genes. Aside from genes for histones and transfer RNA, most of the encoded molecules play an important role in the development, maturation, and regulation of the T cell repertoire and other immunological processes. The MHC regions harbors separate clusters of MHC class I, class II, and class III genes (Fig. 74-2). The classic class I genes are HLA-A, HLA-B, and HLA-C. These genes encode the heavy (α) chain of MHC class I molecules. The class II region contains genes for the α and β chains of the antigen-presenting MHC class II molecules HLA-DR, HLA-DP, and HLA-DQ. The class III region encodes structurally and functionally diverse proteins such as complement components, tumor necrosis factors α and β, and heat shock proteins (HSP70). Other interesting candidates within this region include the gene for myelin oligodendrocyte glycoprotein, a minor structural but immunogenic component of the CNS myelin and so called class I–like genes such as the MHC class I polypeptide-related sequence gene family that regulates the activation and inhibition of natural killer cells.

Figure 74-2 The major histocompatibility complex (MHC) region on chromosome 6p21. To date, the MHC-class II region containing genes for the α and β chains of the antigen-presenting MHC class II molecules human leukocyte antigen (HLA)-DR, HLA-D P, and HLA-DQ remains the only area of the whole genome clearly associated with multiple sclerosis. The most consistent association was found for HLA-DR and HLA-DQ genes.

(Adapted from Janeway CA, et al, eds: Immunobiology. New York: Garland Publishing.)

The most consistent association of multiple sclerosis was found for HLA-DR and HLA-DQ genes. In particular, the HLA-DR15 haplotype (DRB1*1501, DRB5*0101, DQA1*0102, DQB1*0602) showed a strong association with the disease among white persons in northern and central Europe, North America, and Australia, and even a dose effect of HLA-DR15 could be identified in homozygous patients with multiple sclerosis. Because DRB1 and DQB1 genes are in strong linkage disequilibrium, it has been difficult to distinguish the relative contribution from each allelic variant to the susceptibility haplotype. Both additive and independent effects have been reported for the DQA1*0102/DQB1*0602 locus. More recent investigations, however, indicated a primary role of the DRB1*1501 locus independent of DQB1*0602 at least in a white population.⁹

The HLA-DR15 association was comparably weak or lacking in other ethnic groups. Mediterranean populations with a distinct genetic background, such as Sardinians, showed a stronger association with HLA-DR4, and no apparent association with any HLA haplotype could be identified in a southern Chinese population. Much less information is available for risks conferred by class I alleles, although HLA-A3 and HLA-B7 are the first MHC genes that were found to be associated with multiple sclerosis.¹⁰ Their association with multiple sclerosis, however, appears to be much lower than that of the HLA-DR and HLA-DQ alleles. In addition to these risk genes, certain MHC class I and class II allelic variants have been identified as exerting protective properties. Table 74-2 summarizes the reports on risk/protective alleles for multiple sclerosis within the MHC region.

Only limited data are available for the association of clinical features with the HLA-DR/HLA-DQ risk alleles in multiple sclerosis patients. It has been reported that patients expressing the DR15 haplotype have an earlier disease onset, more often have relapsing-remitting multiple sclerosis, are female, and have optic neuritis or spinal involvement as an initial event. DR4-positive patients are described to have a worse clinical outcome or progressive course. However, as mentioned previously, the HLA association studies are heterogeneous with regard to sample size, methodology, ethnic background, and clinical findings. In older studies, the exact MHC class II gene has not even been determined by molecular typing techniques.

HLA-DR and HLA-DQ molecules are by far the strongest genetic risk factors in multiple sclerosis. The mechanisms that are responsible for the genetic association of HLA alleles with multiple sclerosis are, however, incompletely understood. The MHC molecules (1) may fail in contributing to the deletion of autoreactive T cells within the thymus, (2) may preferentially process and present CNS antigens and activate encephalitogenic T cells, or (3) may show an organ- or tissue-dependent expression pattern with a selective detrimental induction in the CNS environment. MHC class I genes may additionally act independently of class II genes in some patients, either through similar mechanisms or by modulation of natural killer cell activity.

FUTURE DIRECTIONS

With the exception of the MHC region on chromosome 6p21, traditional genetic study designs and analytical methods have obviously not been very successful in identifying multiple sclerosis–associated alleles. In future studies, investigators not only must employ more patients but also, of more importance, must use more densely distributed markers throughout the genome such as single nucleotide polymorphisms (SNPs). SNPs are DNA sequence variations that occur when a single nucleotide in the genome sequence is altered. These polymorphisms make up about 90% of all human genetic variation and occur every 100 to 300 bases along the 3 billion–base human genome. SNPs are evolutionarily stable, not changing much from generation to generation, which makes them easier to monitor in population studies. Because of their frequency and stability, SNPs are believed to be more informative than are conventional methods in identifying risk-conferring genes in complex diseases such as multiple sclerosis. Commercial platforms with ever increasing numbers of SNPs become rapidly available, and it has been shown that retyping existing patient cohorts with these tools yields valuable information. Furthermore, it is hoped that additional discovery-oriented techniques such as gene expression profiling by oligonucleotide or complementary DNA microarrays and, at some point, also proteomics will enable clinicians to discern the functional role of susceptibility loci. However, as promising as these approaches may be, they should be combined with better phenotypic characterization of multiple sclerosis patients by neuroimaging, careful clinical characterization, and individual treatment responsiveness. Only the integration of all these observations will eventually allow clinicians to understand the etiology and pathogenesis of multiple sclerosis.