Journal of Neuroimmunology, 32 (1991) 87-94
87
© 1991 Elsevier Science Publishers B.V. 0165-5728/91/$03.50 JNI 01086
Conference Report
Demyelination - - background and mechanisms Kyoto, Japan, 9-11 September 1990 John Kirk The Queen's University of Belfast, Multiple Sclerosis Research Laboratory, Institute of Pathology, Belfast, U.K.
Key words: Demyelination; Multiple sclerosis; Guillain-Barr6 syndrome; Autoimmune process; Viral infection; Experimental allergic encephalomyelitis; Immunological background; Myelin destruction; Immunosuppression
Kyoto city, ancient capital and former seat of the Emperor of Japan was the venue for a recent 2-day symposium on the theme Demyelination," background and mechanisms, organized by Professor Takeshi Yonezawa (Kyoto, Japan) as a satellite to the 11th International Congress of Neuropathology. Participants included neuropathologists, virologists, immunologists, neurologists and other neuroscientists from Japan, the U.S.A., Australia and Europe. Researchers in this area are driven by a common purpose, the identification of the significant mechanisms which lead to demyelination and dysfunction in the human demyelinating diseases typified by multiple sclerosis (MS) and Guillain-Barr6 syndrome (GBS). Knowledge of these mechanisms would be expected to lead to more precisely targeted therapeutic measures and possibly to prevention. While study of riving patients with MS and of MS necropsy tissues is essential, the value of the various animal and tissue culture models cannot be overstated. Both approaches were covered in this symposium in which neuroimmunology was a unifying theme. Discussions
Address for correspondence: John Kirk, BSc, PhD, Department of Pathology, The Queen's University of Belfast, Multiple Sclerosis Research Laboratory, Institute of Pathology, Grosvenor Road, Belfast BT12 6BL, U.K.
covered the dynamics of lesion formation, mechanisms of myelin destruction, inflammation, the possible role of viruses, factors governing disease susceptibility and therapeutic approaches.
CNS and P N S demyelination in man I.V. Allen (Belfast) questioned assumptions about what exactly constitutes a demyelinating disease, pointing out that myelin destruction with relative axonal sparing is a pathological feature common to a variety of acute and chronic conditions with varying clinical and anatomical features. She presented evidence based mainly on enzyme, lipid and immunochemistry to support the idea that MS is a diffuse central nervous system (CNS) disease in which demyelination is the most significant and obvious pathological change but is not the only histological abnormality. She introduced the three most common hypotheses advanced for the aetiology of MS, i.e. the autoimmune, the infective and the combined infective-allergic. She found little evidence to support the first but some (incomplete) evidence to support a viral role as envisaged in the infective or combined infective-allergic hypotheses. J. Kirk (Belfast) described the ultrastructural
88
features of myelin destruction seen in MS, which point to the involvement of soluble factors acting on myelin, prior to myelin phagocytosis in actively demyelinating lesions. The identity (e.g. cytokines, enzymes, radicals, complement, antibody), sources (e.g. lymphocytes, macrophages, astrocytes, serum), and mode of action (enzyme activation, opsonization, etc.) of the significant ones will become clearer as others are eliminated from enquiries. However, myelin destruction is not the earliest alteration in MS for there is ultrastructural, magnetic resonance imaging (MRI) and neuropathological evidence of earlier pathological events (immunological a n d / o r infective) involving blood vessels which should therefore be studied in order to uncover the key pathologic mechanisms. D. Barnes (London) presented the findings of MRI studies of demyelinating diseases, following a period in which new imaging protocols, gadolinium tracer methods and correlative clinical-pathological studies have been carried out in both man and animals and are proving invaluable in understanding the dynamics of lesion formation in MS. The MRI corellates of pathological processes like oedema, gliosis, inflammation, blood-brain barrier leakage and accumulation of lipid debris from myelin breakdown are now being established and within the limits of the specificity of the technique, can now be monitored by MRI as they occur in patients examined at regular intervals. One point which is not often appreciated is that although, on average, patients with MS may suffer clinical relapses about every 2 years, MRI evidence suggests that new patches of damage (lesions or incipient lesions) occur much more frequently (as much as 8 times per year). Another important finding of the MRI work in MS has been the demonstration that cerebral blood vessels in both established lesions and in presumed new lesions during their formation are transiently leaky (typically for about 2 weeks) to the tracer gadolinium-DTPA. There is evidence from both human and animal studies to suggest that such leakage is associated with acute episodes of inflammatory infiltration. The application of MRI methods to patients undergoing specific immunotherapeutic treatments is now capable of providing objective measures of changes occurring in lesions in the brain, a very significant advance on older methods
of assessment which rely mainly on clinical assessment. T. Tabira (Tokyo) reported the clinical and pathological findings in Bal6's concentric sclerosis, a unique variant of MS prevalent in the Philippines. It differs from classical MS in its rapidity of onset, short (80 days) average survival time, in the appearance of its lesions by computerised tomography (CT) and histologically (a distinctive pattern of alternate demyelination and myelin preservation is seen). Inflammatory cuffing is with few exceptions mild, virus inclusions are not seen and the aetiology is unknown. Raine (New York) commented that in his single published case the inflammatory cells had been of the CD8 phenotype, i.e. were class I restricted. C.C.A. Bernard (Bundoora) speculated on the role of intrathecally synthesised autoantibodies to myelin components in the pathogenesis of multiple sclerosis. He postulated that immunoglobulin binding to myelin may result in activation of the myelin neutral protease which uses myelin basic protein (MBP) as substrate. The specificities of MS Igs that stimulate MBP degradation in myelin are unknown but circumstantial evidence suggests that myelin-oligodendrocyte glycoprotein (MOG) could be one target (e.g. monoclonal antibody to M O G results in demyelination both in vivo and in vitro). S. Kusonoki (Tokyo) summarized the evidence for a role of serum antiglycolipid antibodies in the pathogenesis of a demyelinating disease of the peripheral nervous system (PNS), Guillain-Barr6 syndrome. He reported the finding (by enzymelinked immunosorbent assay (ELISA) with confirmation by thin-layer chromatography (TLC) immunostaining) of a variety of serum antiglycolipid antibodies in eight out of 11 GBS patients but not in any controls other than in one case of MS. The most commonly found antibody was anti-GM1 IgM antibody. Recently anti-GM1 antibodies have also been reported in motor-dominant neuropathies and in motor neuron disease so the findings may support the idea that anti-GM1 antibody is associated with motor nerve involvement. Furthermore he reported that in five cases in which the time courses of antiglycolipid antibody activities were followed, antibody titres decreased in association with clinical improvement.
89 Autoimmune processes in demyelination B. Waksman (New York) summarized recent evidence which, he claims, establishes for the first time the presence in MS patients of significant autoimmune reactivity against myelin components. Quoting reports from several laboratories he described the finding in peripheral blood of MS patients of T-cell lines reactive with MBP peptides 84-102 (Hailer et al.) or 87-106 (more precisely 89 or 90-96; McFarland et al.). The HLA restriction was HLA DR2, 4, or 6. In both these studies, cell lines reactive with other peptides of MBP (143-168 or 154-172) were found in blood of controls. In Hafler's HLA DR2 + controis, however, a few lines were found to be reactive to peptides 84-102. A much sharper distinction was found between MS patients and controls in another study (Mokhtarian et al.) in which markers specific for recently activated T-cells (the interleukin-2 (IL-2) receptor) were found only on MBP-reactive cells from MS patients, implying that the MBP-reactive T-cells in Hafler's DR2 + controls were inactive, presumably with little pathogenetic potential. The distinction between MS patients and controls appears to be even more sharply defined when T-cell reactivity against proteohpid (PLP) is measured. If these findings are confirmed and can be shown to be specific for MS and not secondary to the breakdown of myelin in the disease then the advocates of the autoimmune theory of MS pathogenesis may feel they have made significant progress.
Viral infection and demyelination D.E. McFarlin (Bethesda) considered the role of viral infection in producing demyelinating disease. Viral infection can lead to demyelination through a variety of different pathogenic processes. These include lyric and persistent infections of myelin producing cells, an immunopathologic reaction, and the induction of an autoimmune process. The consequences of virus infection depend on a number of variables including the type of virus and the host genetic background. Examples of various pathogenic processes in human de-
myelinating disease and animal models were reviewed. M.C. Dal Canto (Chicago) compared the effectiveness of tolerization in the regulation of experimental allergic encephalomyelitis (EAE) and Theiler's virus-induced demyelinating disease (TMEV). He reported that tolerization to MBP by administration of spleen cells coupled to spinal cord with ethyl carbodiamide prevents the development of adoptively transferred EAE by MBPprimed lymph node cells. He showed that the mechanism of EAE prevention is not dependent on the generation of suppressor cells, but is antigen-specific at the peptide level, dose-dependent and MHC-restricted. He reported that neuroantigen-coupled spleen cells are also capable of preventing relapses in adoptively transferred EAE when injected after the first bout of disease, thereby interfering with a disease process which is already under way. He then asked if the immunopathogenesis of the similar-appearing lesions in the chronic phase of TMEV-induced demyelination was the same as in EAE. It appears not, for tolerization of infected mice with spinal cord-coupled cells has no effect on TMEV-induced demyelinating disease. This confirms other indications that an EAE-like pathogenesis was unlikely. Experiments on the effects of TMEVcoupled cells on the delayed-type hypersensitivity (DTH) response to TMEV in animals immunized with UV-inactivated TMEV showed that the DTH response to the virus was reduced but that the antibody response to the virus was increased. Such increase in antibody was mainly due to increase in IgG1, while a corresponding decrease in IgG2a was seen. IgG2a is driven by the same T-cell subset that drives the DTH response (TH1) while IgG1 is driven by the TH2 cells. S. Sonoda (Kagoshima) described experiments aimed at identifying the humoral and immune effectors responsible for the immunopathological background of human T-cell lymphoma virus-I (HTLV-I)-associated myelopathy/tropical spastic paraparesis (HAM/TSP). Patients in this disease typically have increased titre of anti-HTLV-I antibody and the presence of activated T-cells in peripheral blood and cerebrospinal 'fluid. Isolated CD4 + peripheral blood lymphocytes from patients with HAM/TSP were highly productive of
90 HTLV-I while the unfractionated PBL produced much less virus under the same culture condition. When CD8 cells were added back to the culture of CD4 cells, the CD8 cells recognized HTLV-I-infected CD4 cells and suppressed HTLV-I production to the same level as that of the unfractionated PBL. Similar anti-HTLV-I CD8 effectors were able to be produced from CD8 cells of normal donors whose lymphocytes carried the same HLA haplotypes as those of HAM patients, but not from other HLA haplotype donors. These results suggested that HAM patients might be genetically determined to produce a high response of CD8 effectors and provoke the autoimmune reaction targeting HTLV-I antigens expressed on CD4 cells or the autologous nervous tissues expressing antigens cross-reactive with HTLV-I-infected lymphocytes.
EAE, acute and chronic, and EAN as animal models of demyelinating diseases
H. Lassmann (Vienna) discussed the validity of different models of EAE, the inflammatory demyelinating disease which is induced by autosensitization with brain antigens or by passive transfer of autoreactive T-lymphocytes. Many models have been described, dependent upon the mode of sensitization or passive transfer, the chemical nature of the sensitizing antigen and the genetic background of the immunized animal. EAE models induced by active sensitization, although technically simple, are governed by the complex immunoregulatory mechanisms involved in the sensitization and effector phases. In contrast, passive transfer models (especially if T-cell lines or clones are used) are highly reproducible and predictable and are, thus, especially well suited to study individual aspects of immunoregulation and of the pathogenesis of the disease. Although all models show pronounced brain inflammation, demyelination is highly variable. Pathogenetic studies suggest that a T-lymphocyte response against MBP or PLP is required to induce brain inflammation in EAE while demyelination may be induced by either toxic inflammatory mediators (cytokines, oxygen radicals, etc.) or, more effectively, by the simultaneous presence of circulating demyelinat-
ing antibodies. Although the wide spectrum of different EAE models completely covers the variability in the pathology of human inflammatory demyelinating diseases, no single model covers all aspects simultaneously. Thus the selection of the right model system is of critical importance in research on the pathogenesis and therapy of human inflammatory demyelinating diseases. Y. Matsumoto (Niigata) described attempts to elucidate the immune mechanisms of susceptibility and resistance to EAE. His novel approach was to examine the T-cell repertoire for MBP using thymectomized chimeras that possessed thymuses from an EAE-susceptible (Lew) or an EAE-resistant (BN) strain. It was shown that although the T-cell specificity of these chimeras was skewed towards that of the grafted thymus, the chimeras bearing thymuses from the resistant strain still developed severe EAE. These and other findings suggested that the strain-specific T-cell repertoire itself is not involved in the regulation of EAE susceptibility. Rather, the analysis of the chimeras reconstituted with F1 T-cells and marrow cells from various strains indicates that the major histocompatibility complex (MHC) molecules expressed on accessory cells primarily determine resistance or susceptibility to EAE. Examination of various inbred and congenic rats carrying RT11 or RT1 n revealed that susceptibility to EAE of rats carrying RT11 is heavily influenced by the background genes, whereas resistance to EAE of rats carrying RT1 n is regulated by the MHC molecules expressed on accessory cells without influence of the background genes. M. Pender (Brisbane) described a study of the neuropathology and pathophysiology in different forms of EAE in the Lewis rat. In acute EAE induced by sensitization to MBP or by the passive transfer of MBP-specific lymphocytes, there is inflammation and demyelination in the ventral and dorsal sPinal roots and inflammation and limited demyelination in the spinal cord. More extensive spinal cord demyelination as well as demyelination in the spinal roots and ganglia are observed in acute EAE, induced by inoculation with whole central nervous system (CNS) tissue. In chronic relapsing EAE induced by inoculation with whole CNS tissue and treatment with low-dose cyclosporin A, large plaques of spinal cord demyelina-
91 tion occur in rats with clinical episodes > 28 days after inoculation. In rats with clinical disease < 25 days after inoculation there is prominent demyelination in the spinal roots and ganglia as well as in the spinal cord. In all four forms of EAE, peripheral nervous system remyelination by Schwann cells and CNS remyelination by oligodendrocytes occur during clinical recovery. Electrophysiological studies revealed PNS and CNS nerve conduction abnormalities during clinical episodes, with improvement during remission. T. Saida (Kyoto) described studies aimed at determining the mechanisms of antibody-mediated demyelination, using various anti-galactocerebroside (Gal-C) antibodies (affinity-purified polyclonal and monoclonal IgG and IgM). In vivo PNS demyelination was studied by subperineural micro-injection of antibody a n d / o r cells into Lewis rat sciatic nerves and CNS demyelination in the rabbit eye model by injection into the vitreous body. In congenitally complement-deficient animals and complement-depleted animals injection of anti-Gal-C IgG, IgM, macrophages or large granular cells independently failed to produce demyelination. However, injection of anti-Gal-C IgG together with macrophages induced macrophageassociated demyelination. With high IgG concentration, massive phagocytosis of myelin by macrophages was the dominant feature and with lower concentration IgG, coated pit-associated uptake of myelin fragments into the macrophage cytoplasm was frequently observed. The macrophage-mediated in vivo demyelinative activity of serum correlated with the disease activity of demyelinative neuritis in Gal-C-immunized host animals. This mechanism may also be important, primarily or secondarily in human demyelinating disorders. R. Meyermann (Ttibingen) described the induction and pathology of experimental autoimmune neuritis (EAN), a model of demyelinating polyradiculoneuritis (Guillain-Barrr-Strohl syndrome) in man which is induced by active sensitization of susceptible animals with peripheral nerve tissue or with purified components of PNS myelin. It can also be passively transferred to naive syngeneic animals, by T-lymphocytes isolated from rats with EAN and which react specifically against the P2 protein of the PNS myelin, indicating that EAN is
a T-cell-mediated disease. Although both cellular and humoral immune mechanisms appear to be involved, the mechanism of demyelination in EAN is not known and its extent varies from model to model. That even in acute passively transferred EAE demyelination can be observed, might be explained by the observation that Schwann cells can act as antigen-presenting cells, and, furthermore, can present endogenous myelin proteins. Repeated injections of neuritogenic T cell lines mimic a chronic relapsing inflammatory PNS disease but do not enhance demyelination, whereas demyelination is much more pronounced in actively induced EAN, and is a prominent feature of its chronic variant. Other mechanisms than cellular immune responses are required for selective myelin destruction. Sera of EAN animals have demyelinating effects. In addition, the striking involvement of macrophages in myelin destruction indicates that other factors such as complement, radicals and cytokines are involved in the complexity of PNS myelin breakdown in EAN.
Immunological background of myelin destruction C.S. Raine (New York) described new findings on vascular events leading to autoimmune demyelination. It is widely accepted among proponents of the autoimmune theory of MS pathogenesis that the cellular infiltration leading to demyelination in MS denotes specific recognition between the immune system and its target membrane, myelin, and that it is probably preceded by lymphocyte-endothelial recognition. To explore this, early lesions from cases of MS and mice with EAE, induced by adoptive transfer of MBPspecific lymphocytes, were examined for expression of a battery of adhesion/homing molecules (anti-ICAM-1, anti-MALA-2 (the murine homolog of ICAM-1), anti-LFA-1, HECA-452 and MECA-325 (the latter two being monoclonal antibodies to human and mouse high endothelial venules (HEV) in lymph nodes), and antibodies to class I and class II MHC, CD4 and CD8. Using frozen sections and epoxy-embedded material for immunocytochemistry, he reported that there is up-regulation of several adhesion molecules around blood vessels immediately preceding and
92
during active demyelination. This expression subsides during remissions and increases during relapses. The presence of HEV recognition molecules in MS and EAE lesions implicates the acquisition of lymphoid properties by the CNS during inflammation. Ultrastructural examination of initial attachment events in the EAE model revealed membrane relationships between lymphocytes and endothelial cells which he claimed were unique, further attesting to the specificity of the process. Raine also referred to the findings of a study (Cross et al.) in which the fate of 14C-labelled MBP-specific murine T-cells (mainly CD4 +) which had been activated in vitro to transfer a chronic relapsing form of EAE to syngeneic recipients, was monitored using autoradiography and immunocytochemistry during acute and chronic disease and clinical relapses. Labelled T-cells were first noted in the perivenular areas just prior to onset of clinical disease. In acute disease the labelled cells were found with other cells in perivascular cuffs or attached to the lumenal surface of CNS endothelial cells. They were only rarely seen outside the immediate perivascular area despite the presence of large numbers of unlabelled inflammatory cells in the CNS parenchyma. Labelled cells, which formed only 1-4% of the infiltrates during acute disease, were never seen in the CNS in chronic disease of more than 2 weeks duration, even in areas of extensive myelin pathology. Thus he concluded, MBP-immune Tlymphocytes in this model home to a perivascular location from where they act to recruit a greater number of mainly non-specific, recipient inflammatory cells which proceed to invade the CNS parenchyma and cause demyelination, perhaps by a bystander mechanism. Chronic disease and relapses occur in the apparent absence of donor MBP-immune T-cells and the possibility exists that their place is taken by recruited recipient CNS-antigen-specific T-cells. R.A. Sobel (Boston) described experiments aimed at determining the involvement of cellular adhesion molecules in CNS immune reactions. These molecules, members of the integrin, immunoglobulin gene, cadherin and vascular selectin families mediate adhesion of leukocytes with other leukocytes, with endothelial cells, and with the extracellular matrix in cellular immune responses.
Endothelial cell ICAM-1 surface expression is upregulated early and focally in inflammatory lesions in the CNS of patients with MS and, by binding its ligand, LFA-1, on leukocytes, may promote leukocyte adhesion to and passage through the vessel wall. LFA-1 levels may relate to the selective migration into and functions of specific T-cell subpopulations within inflammatory sites and to immune activation and proliferation in perivascular cuffs. ICAM-1 and LFA-1 on inflammatory cells and glia within MS lesions suggest additional interactions. For example, Creutzfeld cells (glial fibrillary acidic protein (GFAP) positive astrocytes in mitosis) found in acute MS lesions express ICAM-1, implying that they either present antigen or are targets of cytotoxic T-cells. Further studies of the regulation and in situ expression of adhesion molecules should provide insight into mechanisms of CNS immune responses and suggest specific immunotherapy for demyelinating diseases. C.F. Brosnan (New York) demonstrated the potential role of cytokines in lesion development, in particular their relative contributions to inflammation, demyelination and reactive gliosis. Interleukin-1 (IL-1) injection in the rabbit eye model resulted in acute intravascular coagulation and haemorrhage, an early (3 h) and persistent (to 7 days) monocyte response and a wave of polymorphonuclear cells which peaked between 6 and 24 h. Reversible conduction deficits were associated with the peak of the inflammatory response but there was no evidence of demyelination. After treatment with tumour necrosis factor (TNF) for 24 h, mouse organotypic cultures showed intramyelinic oedema localized to periaxonal myelin loops along with some oligodendroglial toxicity. Both T N F and, more markedly, lymphotoxin (LT) exhibited a dose-dependent toxicity to cultures of bovine oligodendrocytes but astrocyte cultures did not show a toxic response under the same circumstances. However, LT, IL-6 and, most effectively, T N F are mitogenic for astrocytes. She concluded that cytokines may play a role in the induction of pathologic changes characteristic of the inflammatory demyelinating diseases and that the relative levels of different cytokines may alter the profile of the cellular inflammatory response. K. Selmaj (New York) provided evidence of the
93 possible involvement of cytokines in the pathogenesis of lesions in multiple sclerosis. Using immunocytochemistry both lymphotoxin (LT) and tumour necrosis factor (TNF) were identified in acute and chronic active MS lesions but were absent from chronic silent lesions. LT was associated with CD3 ÷ lymphocytes and Leu-M5 ÷ microglia at the lesion edge and, to a lesser extent, in adjacent white matter. TNF was associated with astrocytes in all areas of the lesion, and with foamy macrophages in the centre of the active lesion. In acute lesions, immunoreactivity for TNF was found in endothelial cells at the lesion edge. No LT or TNF reactivity was detected in control brain tissue and no increase in reactivity was found in spleen or peripheral blood mononuclear cells (PBMC) of MS patients, suggesting specific reactivity within the CNS. M.B. Graeber (Harvard) described a recently identified population of extraparenchymal cerebral perivascular cells in rat and human brain that bear monocyte/macrophage markers, are easily induced to express MHC class II antigens and thus are candidates for the role of antigen-presenting cell (APC) in the CNS. This finding adds a fourth candidate, APC ceil, to the list already being considered, which comprises microglia, astrocytes and endothelial cells. Unlike microglia which are located within the CNS parenchyma proper, these cells are located adjacent to small and mediumsized CNS blood vessels and are separated from the nervous tissue by a basement membrane. In this anatomic location and with their cell processes alligned with the course of the blood vessels, perivascular cells are well situated to contact infiltrating lymphocytes. The ceils may be identical to the 'perivascular microglia' that have been implicated in antigen presentation in the induction of EAE in vivo. A role for such perivascular cells in human autoimmune disease is thus envisioned.
Therapeutic trials - - immunosuppression
O. Lyon-Caen (Paris) reviewed the effects on MS of various immunotherapeutic regimens which have been introduced on the assumption that MS is an autoimmune disease associated with immune
dysregulation. Corticosteroids improve short-term recovery but have no long-term effect. Intensive use of cyclophosphamide probably has a shortterm effect but some form of maintenance therapy is needed. Azathioprine may be of some benefit, though not in chronic progressive patients, and long-term side effects, probably due to treatment duration, limit its value. Plasma exchange and leukocytopherisis do not clearly benefit MS patients. Interferons (IFN; fl intrathecally; a systemically) have a moderate benefit while "t-IFN makes the disease worse. Cyclosporin has significant side-effects while its efficacy is not demonstrated. Monoclonal antibodies, colchicine and total lymphocyte irradiation (which carried considerable risks) are 'investigational'. Copolymer 1 decreases relapses in early relapsing-remitting MS patients but has not been shown effective in progressive MS. H.L. Weiner recognized the value of nonspecific immunosuppressive agents such as cyclophosphamide in altering the course of progressive MS and, by the use of periodic boosters, in maintaining remission in some patients. However, as he put it, the treatments of today will not be the treatments of the future, since the ultimate goal is to develop non-toxic, immunospecific forms of treatment. He described trials in MS patients and experimental work in animals aimed at achieving this goal. The first area of investigation relates to cell therapy, in which patients are injected with autologous cells cloned from the spinal fluid in an attempt to generate host immune reactivity against putative pathogenic clones. Similar approaches are under consideration using T-cell receptor peptides. The second area of investigation involves antigendriven peripheral immune tolerance to myelin antigens by administration of proteins via the oral route. In this method, the dose of protein necessary to develop tolerance is critical. Studies in animals on both these approaches have shown promise and clinical trials in MS patients are now under way. M.B. Bornstein described the results of clinical trials using copolymer 1 (Cop. 1), the synthetic product of the polymerization of L-glutamic acid, L-lysine and L-tyrosine, which suppresses but does not induce experimental allergic encephalomyelitis (EAE) and is not toxic in animals. Two double-
94
blind, randomized, placebo-controlled pilot trials of Cop. 1 in MS patients have been completed, one in 50 exacerbating-remitting (E-R) and the other in 106 chronic-progressive (C-P) cases. The E-R patients responded at a statistically significant level with frequency of attacks as the major endpoint and degree of disability as the secondary endpoint. C-P patients did not significantly differ from the placebo group for the major endpoint, confirmed progression of 1.0 or 1.5 units on the Kurtzke disability scale, though they did for two secondary endpoints, unconfirmed progression or progression of 0.5 units on the EDSS scale. Side effects noted included mild to moderate irritative reaction at injection sites in most Cop. 1-treated and in some placebo-treated patients, transient vasomotor reactions in 14 Cop. 1 patients and three placebo and urticarial reaction to Cop. 1 in two patients. Current plans are to examine the short- and long-term effects of Cop. 1 in patients with their first attack of acute optic neuritis. In discussing the three papers on therapies, the difficulties of achieving proper 'blinding' were noted. Also, the concept of the 'perfect' therapeutic trial was something which might not be practically achievable in MS research and should not hold us back from doing the best we can.
Conclusions There is room for cautious optimism in this research area which has involved a massive medical and scientific effort over many years, seemingly without major return or breakthrough. However, these foundations were sound and we must acknowledge our debt to the pioneers. It is clear from this symposium that the know-how and technology is now available to establish the immuno-
pathological mechanisms involved in well-characterized immunological and viral animal models and to apply the findings to human disease. Great strides have been made in understanding the processes involved in inflammatory cell recruitment and infiltration into the brain and in defining the potential cellular cites of antigen presentation. The distinct yet interacting roles of the cellular and humoral arms of the immune system in inflammatory demyelinating disease are being dissected. The immunogenetic basis of susceptibility and resistance to EAE is becoming clearer and may have parallels in human disease. The potential roles of antibody, complement enzymes, cytokines, lymphocytes, macrophages and astrocytes in inflammation, gliosis, and demyelination are being catalogued. The evidence that immune attack directed at viral antigens can produce CNS lesions similar to those produced by immune attack on myelin antigens may soon be confirmed by studies in TMEV. Advances in experimental animal research have gone hand in hand with others in human demyelinating disease research. The refinement of MRI is the most notable of these but other technical advances may prove to be equally valuable as they are applied to human tissues, for example the use of immunocytochemistry for detecting cytokines and other effector molecules, the application of in situ hybridization and polymerase chain reaction to the detection of viral genes and other significant nucleic acid species, etc. New evidence of autoimmunity against myelin components in MS may be crucial and needs to be assessed. The extent of peripheral nerve involvement in MS needs to be evaluated. The search for specific immunotherapies for MS is still at an early stage but promising leads are being followed and positive results may eventually be expected.
87
© 1991 Elsevier Science Publishers B.V. 0165-5728/91/$03.50 JNI 01086
Conference Report
Demyelination - - background and mechanisms Kyoto, Japan, 9-11 September 1990 John Kirk The Queen's University of Belfast, Multiple Sclerosis Research Laboratory, Institute of Pathology, Belfast, U.K.
Key words: Demyelination; Multiple sclerosis; Guillain-Barr6 syndrome; Autoimmune process; Viral infection; Experimental allergic encephalomyelitis; Immunological background; Myelin destruction; Immunosuppression
Kyoto city, ancient capital and former seat of the Emperor of Japan was the venue for a recent 2-day symposium on the theme Demyelination," background and mechanisms, organized by Professor Takeshi Yonezawa (Kyoto, Japan) as a satellite to the 11th International Congress of Neuropathology. Participants included neuropathologists, virologists, immunologists, neurologists and other neuroscientists from Japan, the U.S.A., Australia and Europe. Researchers in this area are driven by a common purpose, the identification of the significant mechanisms which lead to demyelination and dysfunction in the human demyelinating diseases typified by multiple sclerosis (MS) and Guillain-Barr6 syndrome (GBS). Knowledge of these mechanisms would be expected to lead to more precisely targeted therapeutic measures and possibly to prevention. While study of riving patients with MS and of MS necropsy tissues is essential, the value of the various animal and tissue culture models cannot be overstated. Both approaches were covered in this symposium in which neuroimmunology was a unifying theme. Discussions
Address for correspondence: John Kirk, BSc, PhD, Department of Pathology, The Queen's University of Belfast, Multiple Sclerosis Research Laboratory, Institute of Pathology, Grosvenor Road, Belfast BT12 6BL, U.K.
covered the dynamics of lesion formation, mechanisms of myelin destruction, inflammation, the possible role of viruses, factors governing disease susceptibility and therapeutic approaches.
CNS and P N S demyelination in man I.V. Allen (Belfast) questioned assumptions about what exactly constitutes a demyelinating disease, pointing out that myelin destruction with relative axonal sparing is a pathological feature common to a variety of acute and chronic conditions with varying clinical and anatomical features. She presented evidence based mainly on enzyme, lipid and immunochemistry to support the idea that MS is a diffuse central nervous system (CNS) disease in which demyelination is the most significant and obvious pathological change but is not the only histological abnormality. She introduced the three most common hypotheses advanced for the aetiology of MS, i.e. the autoimmune, the infective and the combined infective-allergic. She found little evidence to support the first but some (incomplete) evidence to support a viral role as envisaged in the infective or combined infective-allergic hypotheses. J. Kirk (Belfast) described the ultrastructural
88
features of myelin destruction seen in MS, which point to the involvement of soluble factors acting on myelin, prior to myelin phagocytosis in actively demyelinating lesions. The identity (e.g. cytokines, enzymes, radicals, complement, antibody), sources (e.g. lymphocytes, macrophages, astrocytes, serum), and mode of action (enzyme activation, opsonization, etc.) of the significant ones will become clearer as others are eliminated from enquiries. However, myelin destruction is not the earliest alteration in MS for there is ultrastructural, magnetic resonance imaging (MRI) and neuropathological evidence of earlier pathological events (immunological a n d / o r infective) involving blood vessels which should therefore be studied in order to uncover the key pathologic mechanisms. D. Barnes (London) presented the findings of MRI studies of demyelinating diseases, following a period in which new imaging protocols, gadolinium tracer methods and correlative clinical-pathological studies have been carried out in both man and animals and are proving invaluable in understanding the dynamics of lesion formation in MS. The MRI corellates of pathological processes like oedema, gliosis, inflammation, blood-brain barrier leakage and accumulation of lipid debris from myelin breakdown are now being established and within the limits of the specificity of the technique, can now be monitored by MRI as they occur in patients examined at regular intervals. One point which is not often appreciated is that although, on average, patients with MS may suffer clinical relapses about every 2 years, MRI evidence suggests that new patches of damage (lesions or incipient lesions) occur much more frequently (as much as 8 times per year). Another important finding of the MRI work in MS has been the demonstration that cerebral blood vessels in both established lesions and in presumed new lesions during their formation are transiently leaky (typically for about 2 weeks) to the tracer gadolinium-DTPA. There is evidence from both human and animal studies to suggest that such leakage is associated with acute episodes of inflammatory infiltration. The application of MRI methods to patients undergoing specific immunotherapeutic treatments is now capable of providing objective measures of changes occurring in lesions in the brain, a very significant advance on older methods
of assessment which rely mainly on clinical assessment. T. Tabira (Tokyo) reported the clinical and pathological findings in Bal6's concentric sclerosis, a unique variant of MS prevalent in the Philippines. It differs from classical MS in its rapidity of onset, short (80 days) average survival time, in the appearance of its lesions by computerised tomography (CT) and histologically (a distinctive pattern of alternate demyelination and myelin preservation is seen). Inflammatory cuffing is with few exceptions mild, virus inclusions are not seen and the aetiology is unknown. Raine (New York) commented that in his single published case the inflammatory cells had been of the CD8 phenotype, i.e. were class I restricted. C.C.A. Bernard (Bundoora) speculated on the role of intrathecally synthesised autoantibodies to myelin components in the pathogenesis of multiple sclerosis. He postulated that immunoglobulin binding to myelin may result in activation of the myelin neutral protease which uses myelin basic protein (MBP) as substrate. The specificities of MS Igs that stimulate MBP degradation in myelin are unknown but circumstantial evidence suggests that myelin-oligodendrocyte glycoprotein (MOG) could be one target (e.g. monoclonal antibody to M O G results in demyelination both in vivo and in vitro). S. Kusonoki (Tokyo) summarized the evidence for a role of serum antiglycolipid antibodies in the pathogenesis of a demyelinating disease of the peripheral nervous system (PNS), Guillain-Barr6 syndrome. He reported the finding (by enzymelinked immunosorbent assay (ELISA) with confirmation by thin-layer chromatography (TLC) immunostaining) of a variety of serum antiglycolipid antibodies in eight out of 11 GBS patients but not in any controls other than in one case of MS. The most commonly found antibody was anti-GM1 IgM antibody. Recently anti-GM1 antibodies have also been reported in motor-dominant neuropathies and in motor neuron disease so the findings may support the idea that anti-GM1 antibody is associated with motor nerve involvement. Furthermore he reported that in five cases in which the time courses of antiglycolipid antibody activities were followed, antibody titres decreased in association with clinical improvement.
89 Autoimmune processes in demyelination B. Waksman (New York) summarized recent evidence which, he claims, establishes for the first time the presence in MS patients of significant autoimmune reactivity against myelin components. Quoting reports from several laboratories he described the finding in peripheral blood of MS patients of T-cell lines reactive with MBP peptides 84-102 (Hailer et al.) or 87-106 (more precisely 89 or 90-96; McFarland et al.). The HLA restriction was HLA DR2, 4, or 6. In both these studies, cell lines reactive with other peptides of MBP (143-168 or 154-172) were found in blood of controls. In Hafler's HLA DR2 + controis, however, a few lines were found to be reactive to peptides 84-102. A much sharper distinction was found between MS patients and controls in another study (Mokhtarian et al.) in which markers specific for recently activated T-cells (the interleukin-2 (IL-2) receptor) were found only on MBP-reactive cells from MS patients, implying that the MBP-reactive T-cells in Hafler's DR2 + controls were inactive, presumably with little pathogenetic potential. The distinction between MS patients and controls appears to be even more sharply defined when T-cell reactivity against proteohpid (PLP) is measured. If these findings are confirmed and can be shown to be specific for MS and not secondary to the breakdown of myelin in the disease then the advocates of the autoimmune theory of MS pathogenesis may feel they have made significant progress.
Viral infection and demyelination D.E. McFarlin (Bethesda) considered the role of viral infection in producing demyelinating disease. Viral infection can lead to demyelination through a variety of different pathogenic processes. These include lyric and persistent infections of myelin producing cells, an immunopathologic reaction, and the induction of an autoimmune process. The consequences of virus infection depend on a number of variables including the type of virus and the host genetic background. Examples of various pathogenic processes in human de-
myelinating disease and animal models were reviewed. M.C. Dal Canto (Chicago) compared the effectiveness of tolerization in the regulation of experimental allergic encephalomyelitis (EAE) and Theiler's virus-induced demyelinating disease (TMEV). He reported that tolerization to MBP by administration of spleen cells coupled to spinal cord with ethyl carbodiamide prevents the development of adoptively transferred EAE by MBPprimed lymph node cells. He showed that the mechanism of EAE prevention is not dependent on the generation of suppressor cells, but is antigen-specific at the peptide level, dose-dependent and MHC-restricted. He reported that neuroantigen-coupled spleen cells are also capable of preventing relapses in adoptively transferred EAE when injected after the first bout of disease, thereby interfering with a disease process which is already under way. He then asked if the immunopathogenesis of the similar-appearing lesions in the chronic phase of TMEV-induced demyelination was the same as in EAE. It appears not, for tolerization of infected mice with spinal cord-coupled cells has no effect on TMEV-induced demyelinating disease. This confirms other indications that an EAE-like pathogenesis was unlikely. Experiments on the effects of TMEVcoupled cells on the delayed-type hypersensitivity (DTH) response to TMEV in animals immunized with UV-inactivated TMEV showed that the DTH response to the virus was reduced but that the antibody response to the virus was increased. Such increase in antibody was mainly due to increase in IgG1, while a corresponding decrease in IgG2a was seen. IgG2a is driven by the same T-cell subset that drives the DTH response (TH1) while IgG1 is driven by the TH2 cells. S. Sonoda (Kagoshima) described experiments aimed at identifying the humoral and immune effectors responsible for the immunopathological background of human T-cell lymphoma virus-I (HTLV-I)-associated myelopathy/tropical spastic paraparesis (HAM/TSP). Patients in this disease typically have increased titre of anti-HTLV-I antibody and the presence of activated T-cells in peripheral blood and cerebrospinal 'fluid. Isolated CD4 + peripheral blood lymphocytes from patients with HAM/TSP were highly productive of
90 HTLV-I while the unfractionated PBL produced much less virus under the same culture condition. When CD8 cells were added back to the culture of CD4 cells, the CD8 cells recognized HTLV-I-infected CD4 cells and suppressed HTLV-I production to the same level as that of the unfractionated PBL. Similar anti-HTLV-I CD8 effectors were able to be produced from CD8 cells of normal donors whose lymphocytes carried the same HLA haplotypes as those of HAM patients, but not from other HLA haplotype donors. These results suggested that HAM patients might be genetically determined to produce a high response of CD8 effectors and provoke the autoimmune reaction targeting HTLV-I antigens expressed on CD4 cells or the autologous nervous tissues expressing antigens cross-reactive with HTLV-I-infected lymphocytes.
EAE, acute and chronic, and EAN as animal models of demyelinating diseases
H. Lassmann (Vienna) discussed the validity of different models of EAE, the inflammatory demyelinating disease which is induced by autosensitization with brain antigens or by passive transfer of autoreactive T-lymphocytes. Many models have been described, dependent upon the mode of sensitization or passive transfer, the chemical nature of the sensitizing antigen and the genetic background of the immunized animal. EAE models induced by active sensitization, although technically simple, are governed by the complex immunoregulatory mechanisms involved in the sensitization and effector phases. In contrast, passive transfer models (especially if T-cell lines or clones are used) are highly reproducible and predictable and are, thus, especially well suited to study individual aspects of immunoregulation and of the pathogenesis of the disease. Although all models show pronounced brain inflammation, demyelination is highly variable. Pathogenetic studies suggest that a T-lymphocyte response against MBP or PLP is required to induce brain inflammation in EAE while demyelination may be induced by either toxic inflammatory mediators (cytokines, oxygen radicals, etc.) or, more effectively, by the simultaneous presence of circulating demyelinat-
ing antibodies. Although the wide spectrum of different EAE models completely covers the variability in the pathology of human inflammatory demyelinating diseases, no single model covers all aspects simultaneously. Thus the selection of the right model system is of critical importance in research on the pathogenesis and therapy of human inflammatory demyelinating diseases. Y. Matsumoto (Niigata) described attempts to elucidate the immune mechanisms of susceptibility and resistance to EAE. His novel approach was to examine the T-cell repertoire for MBP using thymectomized chimeras that possessed thymuses from an EAE-susceptible (Lew) or an EAE-resistant (BN) strain. It was shown that although the T-cell specificity of these chimeras was skewed towards that of the grafted thymus, the chimeras bearing thymuses from the resistant strain still developed severe EAE. These and other findings suggested that the strain-specific T-cell repertoire itself is not involved in the regulation of EAE susceptibility. Rather, the analysis of the chimeras reconstituted with F1 T-cells and marrow cells from various strains indicates that the major histocompatibility complex (MHC) molecules expressed on accessory cells primarily determine resistance or susceptibility to EAE. Examination of various inbred and congenic rats carrying RT11 or RT1 n revealed that susceptibility to EAE of rats carrying RT11 is heavily influenced by the background genes, whereas resistance to EAE of rats carrying RT1 n is regulated by the MHC molecules expressed on accessory cells without influence of the background genes. M. Pender (Brisbane) described a study of the neuropathology and pathophysiology in different forms of EAE in the Lewis rat. In acute EAE induced by sensitization to MBP or by the passive transfer of MBP-specific lymphocytes, there is inflammation and demyelination in the ventral and dorsal sPinal roots and inflammation and limited demyelination in the spinal cord. More extensive spinal cord demyelination as well as demyelination in the spinal roots and ganglia are observed in acute EAE, induced by inoculation with whole central nervous system (CNS) tissue. In chronic relapsing EAE induced by inoculation with whole CNS tissue and treatment with low-dose cyclosporin A, large plaques of spinal cord demyelina-
91 tion occur in rats with clinical episodes > 28 days after inoculation. In rats with clinical disease < 25 days after inoculation there is prominent demyelination in the spinal roots and ganglia as well as in the spinal cord. In all four forms of EAE, peripheral nervous system remyelination by Schwann cells and CNS remyelination by oligodendrocytes occur during clinical recovery. Electrophysiological studies revealed PNS and CNS nerve conduction abnormalities during clinical episodes, with improvement during remission. T. Saida (Kyoto) described studies aimed at determining the mechanisms of antibody-mediated demyelination, using various anti-galactocerebroside (Gal-C) antibodies (affinity-purified polyclonal and monoclonal IgG and IgM). In vivo PNS demyelination was studied by subperineural micro-injection of antibody a n d / o r cells into Lewis rat sciatic nerves and CNS demyelination in the rabbit eye model by injection into the vitreous body. In congenitally complement-deficient animals and complement-depleted animals injection of anti-Gal-C IgG, IgM, macrophages or large granular cells independently failed to produce demyelination. However, injection of anti-Gal-C IgG together with macrophages induced macrophageassociated demyelination. With high IgG concentration, massive phagocytosis of myelin by macrophages was the dominant feature and with lower concentration IgG, coated pit-associated uptake of myelin fragments into the macrophage cytoplasm was frequently observed. The macrophage-mediated in vivo demyelinative activity of serum correlated with the disease activity of demyelinative neuritis in Gal-C-immunized host animals. This mechanism may also be important, primarily or secondarily in human demyelinating disorders. R. Meyermann (Ttibingen) described the induction and pathology of experimental autoimmune neuritis (EAN), a model of demyelinating polyradiculoneuritis (Guillain-Barrr-Strohl syndrome) in man which is induced by active sensitization of susceptible animals with peripheral nerve tissue or with purified components of PNS myelin. It can also be passively transferred to naive syngeneic animals, by T-lymphocytes isolated from rats with EAN and which react specifically against the P2 protein of the PNS myelin, indicating that EAN is
a T-cell-mediated disease. Although both cellular and humoral immune mechanisms appear to be involved, the mechanism of demyelination in EAN is not known and its extent varies from model to model. That even in acute passively transferred EAE demyelination can be observed, might be explained by the observation that Schwann cells can act as antigen-presenting cells, and, furthermore, can present endogenous myelin proteins. Repeated injections of neuritogenic T cell lines mimic a chronic relapsing inflammatory PNS disease but do not enhance demyelination, whereas demyelination is much more pronounced in actively induced EAN, and is a prominent feature of its chronic variant. Other mechanisms than cellular immune responses are required for selective myelin destruction. Sera of EAN animals have demyelinating effects. In addition, the striking involvement of macrophages in myelin destruction indicates that other factors such as complement, radicals and cytokines are involved in the complexity of PNS myelin breakdown in EAN.
Immunological background of myelin destruction C.S. Raine (New York) described new findings on vascular events leading to autoimmune demyelination. It is widely accepted among proponents of the autoimmune theory of MS pathogenesis that the cellular infiltration leading to demyelination in MS denotes specific recognition between the immune system and its target membrane, myelin, and that it is probably preceded by lymphocyte-endothelial recognition. To explore this, early lesions from cases of MS and mice with EAE, induced by adoptive transfer of MBPspecific lymphocytes, were examined for expression of a battery of adhesion/homing molecules (anti-ICAM-1, anti-MALA-2 (the murine homolog of ICAM-1), anti-LFA-1, HECA-452 and MECA-325 (the latter two being monoclonal antibodies to human and mouse high endothelial venules (HEV) in lymph nodes), and antibodies to class I and class II MHC, CD4 and CD8. Using frozen sections and epoxy-embedded material for immunocytochemistry, he reported that there is up-regulation of several adhesion molecules around blood vessels immediately preceding and
92
during active demyelination. This expression subsides during remissions and increases during relapses. The presence of HEV recognition molecules in MS and EAE lesions implicates the acquisition of lymphoid properties by the CNS during inflammation. Ultrastructural examination of initial attachment events in the EAE model revealed membrane relationships between lymphocytes and endothelial cells which he claimed were unique, further attesting to the specificity of the process. Raine also referred to the findings of a study (Cross et al.) in which the fate of 14C-labelled MBP-specific murine T-cells (mainly CD4 +) which had been activated in vitro to transfer a chronic relapsing form of EAE to syngeneic recipients, was monitored using autoradiography and immunocytochemistry during acute and chronic disease and clinical relapses. Labelled T-cells were first noted in the perivenular areas just prior to onset of clinical disease. In acute disease the labelled cells were found with other cells in perivascular cuffs or attached to the lumenal surface of CNS endothelial cells. They were only rarely seen outside the immediate perivascular area despite the presence of large numbers of unlabelled inflammatory cells in the CNS parenchyma. Labelled cells, which formed only 1-4% of the infiltrates during acute disease, were never seen in the CNS in chronic disease of more than 2 weeks duration, even in areas of extensive myelin pathology. Thus he concluded, MBP-immune Tlymphocytes in this model home to a perivascular location from where they act to recruit a greater number of mainly non-specific, recipient inflammatory cells which proceed to invade the CNS parenchyma and cause demyelination, perhaps by a bystander mechanism. Chronic disease and relapses occur in the apparent absence of donor MBP-immune T-cells and the possibility exists that their place is taken by recruited recipient CNS-antigen-specific T-cells. R.A. Sobel (Boston) described experiments aimed at determining the involvement of cellular adhesion molecules in CNS immune reactions. These molecules, members of the integrin, immunoglobulin gene, cadherin and vascular selectin families mediate adhesion of leukocytes with other leukocytes, with endothelial cells, and with the extracellular matrix in cellular immune responses.
Endothelial cell ICAM-1 surface expression is upregulated early and focally in inflammatory lesions in the CNS of patients with MS and, by binding its ligand, LFA-1, on leukocytes, may promote leukocyte adhesion to and passage through the vessel wall. LFA-1 levels may relate to the selective migration into and functions of specific T-cell subpopulations within inflammatory sites and to immune activation and proliferation in perivascular cuffs. ICAM-1 and LFA-1 on inflammatory cells and glia within MS lesions suggest additional interactions. For example, Creutzfeld cells (glial fibrillary acidic protein (GFAP) positive astrocytes in mitosis) found in acute MS lesions express ICAM-1, implying that they either present antigen or are targets of cytotoxic T-cells. Further studies of the regulation and in situ expression of adhesion molecules should provide insight into mechanisms of CNS immune responses and suggest specific immunotherapy for demyelinating diseases. C.F. Brosnan (New York) demonstrated the potential role of cytokines in lesion development, in particular their relative contributions to inflammation, demyelination and reactive gliosis. Interleukin-1 (IL-1) injection in the rabbit eye model resulted in acute intravascular coagulation and haemorrhage, an early (3 h) and persistent (to 7 days) monocyte response and a wave of polymorphonuclear cells which peaked between 6 and 24 h. Reversible conduction deficits were associated with the peak of the inflammatory response but there was no evidence of demyelination. After treatment with tumour necrosis factor (TNF) for 24 h, mouse organotypic cultures showed intramyelinic oedema localized to periaxonal myelin loops along with some oligodendroglial toxicity. Both T N F and, more markedly, lymphotoxin (LT) exhibited a dose-dependent toxicity to cultures of bovine oligodendrocytes but astrocyte cultures did not show a toxic response under the same circumstances. However, LT, IL-6 and, most effectively, T N F are mitogenic for astrocytes. She concluded that cytokines may play a role in the induction of pathologic changes characteristic of the inflammatory demyelinating diseases and that the relative levels of different cytokines may alter the profile of the cellular inflammatory response. K. Selmaj (New York) provided evidence of the
93 possible involvement of cytokines in the pathogenesis of lesions in multiple sclerosis. Using immunocytochemistry both lymphotoxin (LT) and tumour necrosis factor (TNF) were identified in acute and chronic active MS lesions but were absent from chronic silent lesions. LT was associated with CD3 ÷ lymphocytes and Leu-M5 ÷ microglia at the lesion edge and, to a lesser extent, in adjacent white matter. TNF was associated with astrocytes in all areas of the lesion, and with foamy macrophages in the centre of the active lesion. In acute lesions, immunoreactivity for TNF was found in endothelial cells at the lesion edge. No LT or TNF reactivity was detected in control brain tissue and no increase in reactivity was found in spleen or peripheral blood mononuclear cells (PBMC) of MS patients, suggesting specific reactivity within the CNS. M.B. Graeber (Harvard) described a recently identified population of extraparenchymal cerebral perivascular cells in rat and human brain that bear monocyte/macrophage markers, are easily induced to express MHC class II antigens and thus are candidates for the role of antigen-presenting cell (APC) in the CNS. This finding adds a fourth candidate, APC ceil, to the list already being considered, which comprises microglia, astrocytes and endothelial cells. Unlike microglia which are located within the CNS parenchyma proper, these cells are located adjacent to small and mediumsized CNS blood vessels and are separated from the nervous tissue by a basement membrane. In this anatomic location and with their cell processes alligned with the course of the blood vessels, perivascular cells are well situated to contact infiltrating lymphocytes. The ceils may be identical to the 'perivascular microglia' that have been implicated in antigen presentation in the induction of EAE in vivo. A role for such perivascular cells in human autoimmune disease is thus envisioned.
Therapeutic trials - - immunosuppression
O. Lyon-Caen (Paris) reviewed the effects on MS of various immunotherapeutic regimens which have been introduced on the assumption that MS is an autoimmune disease associated with immune
dysregulation. Corticosteroids improve short-term recovery but have no long-term effect. Intensive use of cyclophosphamide probably has a shortterm effect but some form of maintenance therapy is needed. Azathioprine may be of some benefit, though not in chronic progressive patients, and long-term side effects, probably due to treatment duration, limit its value. Plasma exchange and leukocytopherisis do not clearly benefit MS patients. Interferons (IFN; fl intrathecally; a systemically) have a moderate benefit while "t-IFN makes the disease worse. Cyclosporin has significant side-effects while its efficacy is not demonstrated. Monoclonal antibodies, colchicine and total lymphocyte irradiation (which carried considerable risks) are 'investigational'. Copolymer 1 decreases relapses in early relapsing-remitting MS patients but has not been shown effective in progressive MS. H.L. Weiner recognized the value of nonspecific immunosuppressive agents such as cyclophosphamide in altering the course of progressive MS and, by the use of periodic boosters, in maintaining remission in some patients. However, as he put it, the treatments of today will not be the treatments of the future, since the ultimate goal is to develop non-toxic, immunospecific forms of treatment. He described trials in MS patients and experimental work in animals aimed at achieving this goal. The first area of investigation relates to cell therapy, in which patients are injected with autologous cells cloned from the spinal fluid in an attempt to generate host immune reactivity against putative pathogenic clones. Similar approaches are under consideration using T-cell receptor peptides. The second area of investigation involves antigendriven peripheral immune tolerance to myelin antigens by administration of proteins via the oral route. In this method, the dose of protein necessary to develop tolerance is critical. Studies in animals on both these approaches have shown promise and clinical trials in MS patients are now under way. M.B. Bornstein described the results of clinical trials using copolymer 1 (Cop. 1), the synthetic product of the polymerization of L-glutamic acid, L-lysine and L-tyrosine, which suppresses but does not induce experimental allergic encephalomyelitis (EAE) and is not toxic in animals. Two double-
94
blind, randomized, placebo-controlled pilot trials of Cop. 1 in MS patients have been completed, one in 50 exacerbating-remitting (E-R) and the other in 106 chronic-progressive (C-P) cases. The E-R patients responded at a statistically significant level with frequency of attacks as the major endpoint and degree of disability as the secondary endpoint. C-P patients did not significantly differ from the placebo group for the major endpoint, confirmed progression of 1.0 or 1.5 units on the Kurtzke disability scale, though they did for two secondary endpoints, unconfirmed progression or progression of 0.5 units on the EDSS scale. Side effects noted included mild to moderate irritative reaction at injection sites in most Cop. 1-treated and in some placebo-treated patients, transient vasomotor reactions in 14 Cop. 1 patients and three placebo and urticarial reaction to Cop. 1 in two patients. Current plans are to examine the short- and long-term effects of Cop. 1 in patients with their first attack of acute optic neuritis. In discussing the three papers on therapies, the difficulties of achieving proper 'blinding' were noted. Also, the concept of the 'perfect' therapeutic trial was something which might not be practically achievable in MS research and should not hold us back from doing the best we can.
Conclusions There is room for cautious optimism in this research area which has involved a massive medical and scientific effort over many years, seemingly without major return or breakthrough. However, these foundations were sound and we must acknowledge our debt to the pioneers. It is clear from this symposium that the know-how and technology is now available to establish the immuno-
pathological mechanisms involved in well-characterized immunological and viral animal models and to apply the findings to human disease. Great strides have been made in understanding the processes involved in inflammatory cell recruitment and infiltration into the brain and in defining the potential cellular cites of antigen presentation. The distinct yet interacting roles of the cellular and humoral arms of the immune system in inflammatory demyelinating disease are being dissected. The immunogenetic basis of susceptibility and resistance to EAE is becoming clearer and may have parallels in human disease. The potential roles of antibody, complement enzymes, cytokines, lymphocytes, macrophages and astrocytes in inflammation, gliosis, and demyelination are being catalogued. The evidence that immune attack directed at viral antigens can produce CNS lesions similar to those produced by immune attack on myelin antigens may soon be confirmed by studies in TMEV. Advances in experimental animal research have gone hand in hand with others in human demyelinating disease research. The refinement of MRI is the most notable of these but other technical advances may prove to be equally valuable as they are applied to human tissues, for example the use of immunocytochemistry for detecting cytokines and other effector molecules, the application of in situ hybridization and polymerase chain reaction to the detection of viral genes and other significant nucleic acid species, etc. New evidence of autoimmunity against myelin components in MS may be crucial and needs to be assessed. The extent of peripheral nerve involvement in MS needs to be evaluated. The search for specific immunotherapies for MS is still at an early stage but promising leads are being followed and positive results may eventually be expected.
