Although there appears to be an "autoimmune" attack against myelin and myelin-forming cells in the brain and spinal cord, multiple sclerosis cannot be called a true autoimmune disease. T cell and antibody reactivity have been tested against numerous virus and brain antigens, but no target antigen has been clearly demonstrated. The antigen-induced animal model, experimental allergic encephalomyelitis, does not appear spontaneously in wild mice. HLA types are associated, but the mechanism is unclear. There are surprisingly few links to autoimmune disease, except Crohn disease and possibly thyroid disease. Systemic lupus erythematosus is underrepresented in multiple sclerosis and is linked to opposite responses to type I interferons (Javed and Reder 2006).
Specific antigenic targets for inflammation in multiple sclerosis. Candidate central nervous system antigens and targets include:
The antibody response to central nervous system antigens varies between patients. Anti-myelin basic protein responses are weak in multiple sclerosis, differing from the strong responses in animal models. However, pro-inflammatory cells recognizing myelin basic protein are increased when low concentrations of myelin basic protein are used to detect high avidity human T cell clones (Bielekova et al 2004). Anti-proteolipid antibodies in CSF are more common in women than men, in patients with later onset of multiple sclerosis, in patients without a family history of multiple sclerosis, and in those who have low levels of CSF immunoglobulin and oligoclonal bands (Warren et al 1994). Antibodies to myelin oligodendrocyte protein are debatably elevated in all forms of multiple sclerosis (and other inflammatory brain diseases). Antibodies to myelin basic protein are low in early multiple sclerosis and increase over time (Reindl et al 1999), but detection is erratic between laboratories. Even if antibodies to brain antigens do not cause multiple sclerosis, they could modify disease course.
Arguments are made against the presence of a “multiple sclerosis antigen.” For instance, 1 in 220 people vaccinated with the Semple rabies vaccine—which contains central nervous system tissue—develop autoimmune encephalitis (similar to EAE). Patients susceptible to this encephalitis, however, have a human leukocyte antigen (HLA) makeup that is distinct from multiple sclerosis patients (Piyasirisilp et al 1999).
The lack of a causative antigen suggests that fundamental control of immune responses may be abnormal and that oligodendroglia are innocent bystanders damaged by unregulated inflammation. Activated lymphocytes and monocytes might enter the central nervous system because of nonspecific adhesion to endothelial cells, become activated within the central nervous system, stay longer during trafficking through the central nervous system, and escape from the normal CNS suppression of the immune response. Putative antigen-specific responses are described below.
Non-antigen-specific immunity for inflammation in multiple sclerosis. Etiologies that do not invoke specific target antigens are possible in multiple sclerosis.
Viruses. Through direct damage to oligodendroglia, by retrovirus incorporation into oligodendroglia and T cells, and from immune reactivity to shared determinants between oligodendroglia and viruses. The role of human herpes virus-6 and endogenous retroviruses awaits confirmation in multiple sclerosis. Human endogenous retroviruses, HERV, which make up 10% to 30% of the human genome correlate with a more progressive course. However, detection of these viruses is possibly a byproduct of immune activation of viruses and not the cause of the disease. Activated astrocytes produce retrovirus-encoded syncytin, which is toxic to oligodendrocytes.
Antibodies to Epstein-Barr virus correlate with brain atrophy and are elevated early in the course of multiple sclerosis. This may simply reflect multiple sclerosis-characteristic high titers to many antigens and many viruses, possible because HLA-DR2 is over-represented in multiple sclerosis and because DR2-positive people have higher antibody titers to Epstein-Barr virus, measles, and rubella (Compston et al 1986). Anti-Epstein-Barr virus antibodies could arise from persistent infection of astrocytes or B cells, causing costimulatory molecule expression, IL-6 secretion, and immune activation. Epstein-Barr virus infects B cells and could generate an autoreactive B cell population resistant to apoptosis and immune control.
Antibodies to cytomegalovirus, in contrast, correlate with better outcome (Zivadinov et al 2006). Varicella-zoster virus DNA increases briefly in mononuclear cells during relapses, but this virus does not increase the risk of multiple sclerosis. Report of varicella-zoster virus particles in multiple sclerosis brains has not been confirmed (Burgoon et al 2009).
In children, Epstein-Barr virus NA-1 seropositivity increases the risk of multiple sclerosis 3.8-fold. Cytomegalovirus positive serum confers a lower risk of multiple sclerosis in children 0.27-fold (Waubant et al 2011).
Bacteria and chlamydia. Through cross-reactive antigens, superantigen activation of pathogenic T cells, responses to induced heat shock proteins (all trigger cytokine release), and release of bacterial toxins, possibly from posterior sinuses and submucosa (Gay 2007). Conversely, parasite infestation could be protective.
Oligodendroglia. Defective function or repair.
Diet. Affects immunity through oral tolerance and shapes the microbiome. Diet can modify macrophage function, membrane composition of immune cells, and prostaglandin synthesis.
Genetic. Predisposition to respond to brain antigens, altered control of the immune response to brain antigens, lack of neurotrophic proteins, or poor ability to repair CNS damage.
Other mechanisms. Toxins, microchimerism of circulating blood cells, and endocrine, catecholamine, and stress interrelations with immunity have been proposed.
In the 1950s, anticoagulants failed to significantly impact the course of multiple sclerosis based on a theory that CNS microvessels had poor blood flow. Recent use of venous stenting to reverse putative cerebral venous outflow problems (CCSVI) has not been beneficial in controlled studies, although anecdotes of benefit are common. Tens of millions of dollars in research money and medical costs, huge amounts of investigators’ intellectual energy, and misplaced hope by patients are being directed at this questionable therapy.