More MS news articles for Sep 2001

Immunotherapy of multiple sclerosis: Where are we? Where should we go?

Volume 2 No 9 - September 2001
Nature Immunology 2, 785 - 788 (2001) © Nature America, Inc.

Roland Martin1, Claus-Steffen Stürzebecher (2) & Henry F. McFarland (1)

(1) Neuroimmunology Branch, NINDS, National Institutes of Health, Building 10, Room 5B-16, 10 Center DR MSC 1400, Bethesda, MD 20892-1400, USA.
(2) Clinical Development CNS, Schering AG, Berlin 13342, Germany. (

Differences in multiple sclerosis patient's disease and their responses to standard drugs indicate that today's therapies need to be more individualized. It is proposed that gene expression profiling in conjunction with magnetic resonance imaging be used to optimize future treatment approaches.

The treatment of autoimmune diseases is still in its infancy, and corticosteroids and immunosuppression remain the mainstay therapies. More subtle strategies of immunomodulation—such as the use of interferon b  (IFN-b) and glatiramer acetate (GA) in the treatment of multiple sclerosis (MS)—are well accepted, and the use of engineered antibodies or receptors to tumor necrosis factor-a (TNF-a) in the treatment of rheumatoid arthritis (RA) have been introduced as new therapies. However, the heterogeneity of autoimmune diseases, as well as characteristics of the target tissue, pose challenges to induce more effective immune intervention. Recent advances in our understanding of autoimmune disease pathogenesis, in imaging techniques and in the use of other biomarkers, require the reassessment of conventional therapeutic approaches and the development of new ones. Using MS as an example, we describe here the current state of immunotherapy, highlight important questions and stress where we should focus our efforts in the future.

The problem

MS reflects all the potentials for both success and failure in immunotherapy. With respect to successes, MS is probably one of the best understood autoimmune diseases and several treatments are now approved (1). Regarding failures, many approaches found promising in animal models have failed in clinical testing, indicating that MS is considerably more complex than its rodent model, experimental allergic encephalomyelitis (EAE). As a prototype autoimmune disease, MS also reflects remarkable disease heterogeneity— both clinically and pathologically—which, in turn, poses substantial challenges for the design and testing of new therapies. We will briefly discuss MS as an autoimmune process, examine disease heterogeneity and look at possible approaches to find effective future therapies.

The pathogenetic and heterogeneous nature of MS

Although MS presents with a broad range of different symptomatologies and courses—including relapsing-remitting, secondary progressive and primary progressive—it has, nevertheless, been considered as one disease (2, 3). This current taxonomy poses many problems not only for identifying effective treatments, but for all aspects of research into the cause and pathogenesis of MS. Disentangling disease heterogeneity requires the identification of the contributing factors and their relative importance in a given patient or subgroups of patients.

MS is considered to be a T cell–mediated autoimmune disease (4), with various degrees of myelin and axonal damage in the central nervous system (CNS) (5). Its immunological basis is supported by its association with immunologically relevant genes, particularly of the major histocompatibility complex (MHC), and response to immunosuppression. The strongest evidence, however, stems from animal models such as EAE or virus-induced encephalomyelitis (6), which show that T cells that are reactive against self-proteins cause autoimmune demyelinating diseases.

Studies of EAE, as well as the cellular immune response in MS patients and the pathology of MS tissue, have led to the following pathogenetic scenario (Fig. 1). Not all autoreactive T cells are deleted during thymic selection, which means we all harbor T cells with the potential to recognize autoantigens, including myelin proteins (7). Under the right conditions, such as viral infections, these cells may become activated and inflict harm. Upon activation, autoreactive T cells migrate into the CNS. Upon breaching the blood-brain-barrier (BBB), these autoreactive CD4+ T cells initiate the local proinflammatory cascade. Eventually, a variety of effector mechanisms—including antibody-mediated cytotoxicity, oxygen and nitrogen radicals, toxic cytokines and apoptosis-mediating molecules that damage oligodendrocytes, myelin sheaths and occasionally, at this stage, axons—are induced (Fig. 1).

Figure 1. Schematic depiction of the pathogenetic steps that lead to tissue damage in MS. The major steps of the pathogenetic cascade are highlighted: (1) activation of autoreactive T cells in the periphery; (2) transmigration of proinflammatory T cells and monocytes through the BBB; (3) amplification of local inflammation and activation of resident APCs, such as microglia; (4) effector stage of the disease, with damage of oligodendrocytes, myelin sheath and axons. A possible further stage, the resolution of lesions by regulatory mechanisms and remyelination, is not highlighted separately.

Pathological analyses of MS tissue have indicated that the effector mechanisms may vary among patients (8). These observations teach us valuable lessons. Inflammation dominates the picture in most patients, but seems to be quickly resolved by immunoregulatory mechanisms. The acute inflammatory stage is reflected clinically by exacerbations during the relapsing phase of the disease and can be seen on magnetic resonance imaging (MRI) scans as disruption of the BBB with extravasation of contrast-enhancing agents (for example, gadolinium-DTPA, Gd) (Fig. 2). Progression of disability occurs through two separate mechanisms: failure to recover from an attack and slow progression-independent exacerbations. The underlying mechanisms of these progression patterns is likely to differ. The first seems to be related to the extent of destruction and failure to repair after acute lesions. The second is poorly understood, but is probably related to tissue destruction that is independent of an inflammatory response. One concept of the progression of MS is that—once several bouts of inflammation have been inflicted and damaged oligodendrocytes fail to recover—inflammation decreases and an insidious, secondary chronic progressive course with a reduction in Gd-enhancing lesions, loss of tissue, glial scarring, reduction of axons and increasing neurological deficit sets in. Hence, disease is no longer mainly driven by inflammation, whereas the lack of remyelination and loss of glial and neural components now predominate.

Figure 2. Longitudinal course of the fluctuations of contrast-enhancing MRI lesions in a patient who responded to IFN-b treatment. The number of gadolinium-enhancing lesions detected during monthly MRI scans done over several years of follow-up examinations are shown. The initiation of IFN-b therapy is indicated; it led to a dramatic loss of inflammatory MRI lesions in the brain of this patient, who qualified as a drug-responder.

How does pathological and immunological heterogeneity translate into clinical disease heterogeneity, that is, a benign course with some inflammation and tissue damage; a strong proinflammatory immune response with frequent and severe destruction that leads to a progressive disease; or even a course that is characterized by degenerative aspects from onset? We cannot answer these questions yet. However, the pathological, MRI, immunological and clinical evidence all suggest that an aberrant immune response, disturbed immunoregulation and the vulnerability of the target tissue—either genetically determined or due to a persistent virus—all contribute to the heterogeneity of MS.

Our inability to precisely characterize patients represents an important problem for all immunological and genetic studies. Although MRI is of some help in examining part of the heterogeneity (9) (Fig. 2), additional biomarkers—which capture the state of immune dysfunction and tissue damage to distinguish different subtypes and stages of the disease—are lacking. Both genetic background and environmental triggers are important pathogenetic factors for MS. Although our understanding of this is far from complete, it is clear that autoimmune diseases such as RA or MS have a strong genetic predisposition. Unlike disorders with simple Mendelian inheritance, such as Huntington's chorea, autoimmune diseases are caused by a complex genetic trait to which a larger number of genes contribute in a quantitative manner. Human leukocyte antigen (HLA) genes, for example, HLA-DRB1*1501, DRB5*0101 and DQB1*0602, have consistently been identified as conferring MS susceptibility (4). Other susceptibility alleles seem weakly associated and difficult to identify, but a number of these genes are shared by multiple autoimmune disorders (10). Many aspects of immune function may be affected by these genetic factors and, depending on the individual set of susceptibility alleles, the secretion of proinflammatory cytokines may be enhanced in one and apoptosis or immune regulation disturbed in another. The contribution of many genes to the quantitative trait implies that individual patients express different subsets of susceptibility genes and their composition will greatly influence the disease phenotype. If the "genetic dose" surpasses a certain threshold, disease develops in a stochastic manner, but is probably modulated by environment, with susceptibility being related to the number of genes that contribute to the quantitative trait.

Genes that affect the target tissue also probably influence disease initiation and involvement of a specific organ. In MS, the different pathological and clinical patterns suggest that CNS damage is, in some patients, primarily driven by the vulnerability of oligodendrocytes and/or axons. Hence, whereas MS patients may have the same susceptibility genes as other patients suffering from a different autoimmune disease, tissue-specific genetic factors affect which organ or tissue is affected in the disease. Organ specificity may also be related to the tropism of an infectious organism or by molecular mimicry between self-antigens and proteins expressed by the organism. In addition, persistent viral infection and reactivation in the CNS—as in the case of human herpes virus-6 (HHV-6)—may serve as the local factor that disturbs tissue function and causes inflammation.

Approved therapies

MS is among the few autoimmune diseases for which immunomodulatory therapies, IFN-b and GA, have been approved. IFN-b was originally introduced as an antiviral agent; GA was designed to induce EAE, but was then found to be therapeutically active. Both agents are only moderately effective: they have reduced disease exacerbations by approx. 30% or delayed disease progression or onset in large phase III trials (11, 12). The mechanism of action of either agent is incompletely known. IFN-b induces a larger number of cytokines, including interleukin 10 (IL-10), and profoundly influences BBB opening via interference with cell adhesion, migration and matrix metalloproteinase activity. GA, at least in vitro, blocks antigen presentation but, more importantly, induces both a shift from T helper type 1 (TH1) to TH2 cytokines and GA-specific TH2 cells, which cross-react with myelin components and thus mediate bystander suppression (13).

Although the mechanism of IFN-b mentioned above fits well with our current concepts, data from some studies showed that IFN-b also up-regulates TH1-associated cytokines and receptors (for example, the IL-12 receptor b2 chain), chemokines and their receptors (for example, IP-10 and its receptor CCR5) and other proinflammatory genes (14). What do these observations mean? Proinflammatory mediators may have beneficial effects. In addition to their pathological potential, proflammatory mediators may have beneficial effects. Alternatively, the balance of pro- and anti-inflammatory pathways influenced by IFN-b results in an anti-inflammatory response overall or in desensitization of the immune response to damaging stimuli such as viral infections. One could also hypothesize that the profound effects on the BBB may domiante over all other mechanisms.

Failed attempts

Probably the most puzzling example of failed therapies is TNF-a, which was identified as a candidate cytokine involved in tissue damage not only in RA, but also in MS. Distinct from the use of antibodies to TNF-a or soluble TNF receptors to treat RA, which had beneficial effects, attempts to block TNF-a in the treatment of MS worsened disease (15). The reasons for this discrepancy are unclear, but some proinflammatory mediators may also have beneficial effects in a specific organ, such as aiding the resolution of the immune response or even fostering tissue repair and remyelination (16). In addition, multiple cell types and highly redundant networks of cytokines, chemokines and other molecules are involved in migration and homing, as well as effector functions. This redundancy has become obvious with gene manipulation experiments (17). With respect to protection from EAE induction, genes thought to be crucial for autoimmune diseases produced hardly any phenotypic changes. The pleiotropic and overlapping effects of many mediators on various differentiation and effector pathways make it difficult to predict their therapeutic effect.

Despite therapeutic promise, other candidates such as transforming growth factor a (TGF-a) or linomide have proved too toxic when tested in MS patients. Thus, the difficulty of predicting either therapeutic potential or toxicity in human beings makes attempts at affecting cytokine pathways difficult.

Targeting specific autoreactive T cells

Global immunosuppression has never been very appealing to immunologists. Instead, we aim to target the immunological "defect" in autoimmunity as precisely as possible. One such way is to block specific autoreactive cells or antibodies. If we knew which autoreactive T cells contributed to disease, specific vaccination approaches could be designed to eliminate these cells. Early clinical trials towards this end are underway; both T cell receptor (TCR) peptides derived from the complementarity-determining region 2 (CDR2) and CDR3 regions of autoreactive T cells and immunizations with whole inactivated myelin basic protein (MBP)-specific T cells have been studied experimentally or are in clinical testing. Despite promising reports with EAE, we consider it unlikely that these will be effective in humans. This caution stems from the fact that TCR usage of autoreactive T cells in the outbred human population is diverse, and only some individuals show restricted TCR expression, as observed in many EAE models. Consequently, such therapies would require extensive studies of the TCR repertoire in every individual, which is clearly not practical. Ongoing studies will show whether these assumptions are correct. Other strategies, such as the intravenous injection of putative autoantigens to induce activation-induced cell death, are compromised by different problems. The lack of knowledge of "the" autoantigen/s and the need to treat at a time when the autoreactive T cells are in a state of activation carries the risk of exacerbating disease.

Induction of tolerance and immunoregulation

Various approaches to restoring peripheral tolerance or immunoregulation have been successful in EAE. These include oral or nasal administration of autoantigen, injection of fixed antigen-presenting cells (APCs) pulsed with autoantigenic peptide, induction of CD25+ regulatory cells, abrogation of immune cells via irradiation and chemotherapy with subsequent infusion of hematopoietic stem cells (18, 19). Besides the stem cell approach, the common themes are a shift from a proinflammatory (TH1) to an anti-inflammatory (TH2 or TH3) state, anergy and the induction of regulatory cells that downmodulate disease activity via IL-10, TGF- or TH2 cytokines. A particular attractive strategy has been to use autoantigenic peptides with modifications in TCR contact positions, that is, altered peptide ligands (APLs). APLs can mediate anergy, TCR antagonism or, most interestingly, bystander suppression (20). The latter mechanism refers to the induction of an APL-specific TH2-like cell population that cross-reacts with the native autoantigen and thus dampens immune responsiveness whenever autoantigen is released. Recently, an APL of the immunodominant MBP peptide (amino acids 83–99) was tested in phase II trials (21, 22), and several interesting observations emerged. Unexpectedly, the APL induced allergic reactions in a substantial number of MS patients. In addition, at a high dose, some patients showed disease exacerbations that were mediated by APL-specific TH1 cells with cross-reactivity with the native MBP peptide (21). In contrast, a lower dose showed a trend towards clinical benefit, probably via a TH2 shift (22). Although these studies show that our basic concepts about disease induction by specific autoantigens are probably correct, they also highlight that the correct dose of APL and its route of administration need further investigation.

GA also acts, at least in part, through bystander suppression; due to its random composition of four amino acids, it offers greater safety than a defined APL. However, this may be at the expense of lower efficacy than is theoretically achievable with an APL.

Similar considerations apply to oral tolerance. Orally administered whole bovine myelin was not effective in a phase III trial in MS. Whether this was due to choosing the wrong dose or antigen is currently not clear. Stem cell transplants represent the most radical approach to restoring peripheral tolerance. However, toxicity is still relatively high, and it is unclear whether autologous stem cell transplants will prevent re-emerging autoimmunity or at what stage of disease, administration of treatment will be most beneficial.

Will one treatment fit them all?

One treatment is unlikely to fit all. Consequently, our current practice of treating MS with a single drug is probably too simplistic. It is obvious that, depending on their major effect, most treatments will have an impact on some of the pathogenetic steps in MS, but have little effect on others. Thus, and perhaps not surprisingly, the efficacy of the established therapies is only modest. In addition, current treatments are primarily aimed at blocking the autoimmune process, which is reasonable during early relapsing-remitting MS, when inflammation is the primary driving force of the disease process. However, after repeated bouts of inflammation with demyelination and axonal transections, repair through remyelination will be, at best, of limited benefit. Hence, during the advanced stages of MS, we need to develop entirely different therapies aimed at repair, but at the same time inflammation also has to be controlled. These considerations also apply to other autoimmune diseases, such as diabetes and vitiligo.

One solution to this problem might be to treat patients early to maximize the time before irreversible tissue injury occurs. The initiation of IFN-b treatment during very early phases of the inflammatory process delays diseases onset by a few months (23). However, in cases where degenerative aspects—such as loss of oligodendrocytes—appear early, blocking inflammation will have only marginal impact.

Where should we go?

Despite substantial progress, treatments of MS and other autoimmune diseases are only moderately effective. Because we anticipate that the treatment of autoimmune diseases with single drugs or biological approaches will in the future be complemented, or even replaced, by combination therapies, we believe that a number of areas deserve particular attention. We need a better understanding of every aspect of the complexities of the disease process, that is, genetic background, environmental triggers, immune reactivity, vulnerability of the target tissue and pathological and clinical heterogeneity. We need tools to dissect these processes in the individual patient. These include gene expression profiling with cDNA microarrays and proteomics, linking this information to candidate genes and staging the disease process as accurately as possible with in vivo imaging by MRI. Other biomarkers that fulfill the criteria of surrogates and will predict disease severity, course and type of organ damage will be essential for rational therapies. It is imperative that we also obtain a better understanding of how existing and future therapies work. Often we begin to learn how a drug acts only long after its introduction into clinical practice. The use of biomarkers and imaging to characterize the disease in individual patients should be developed so that the most appropriate treatment strategies are used. Drug responders and nonresponders have to be detected early, ideally even before treatment is initiated. Our recent observations using cDNA microarrays to determine gene expression in IFN-b responders and nonresponders (C. S. Stürzebecher et al., unpublished data) show that this may indeed be feasible (Fig. 3). Finally, similar to cancer and many infections, we anticipate that treatment of autoimmune diseases with single drugs or biological approaches will soon be the exception and replaced by combination therapies that include immunomodulation, elimination of infectious triggers and tissue repair.

Figure 3. Identification of responders and nonresponders to immunomodulatory treatment with IFN-b. MRI, cDNA microarrays, quantitative polymerase chain reactions and, later, proteomics were used. The schematic shows how the identifications were done. T, treatment; B, baseline.

The treatment of autoimmune diseases such as MS has progressed remarkably over the past decade. Wisely using the scientific tools that are now available should lead to even more rapid and dramatic progress in the development of treatments over the coming years.

1. Bielekova, B. & Martin, R. Curr. Treat. Options Neurol. 1, 201-219 (1999). | PubMed |
2. McFarlin, D. E. & McFarland, H. F. New Engl. J. Med. 307, 1183-1188 (1982). | PubMed | ISI |
3. McFarlin, D. E. & McFarland, H. F. New Engl. J. Med. 307, 1246-1251 (1982). | PubMed | ISI |
4. Martin, R., McFarland, H. F. & McFarlin, D. E. Annu. Rev. Immunol. 10, 153-187 (1992). | PubMed | ISI |
5. Prineas, J. W. in Handbook of Clinical Neurology, Demyelinating Diseases (eds Vinken, P. J., Bruyn, G. W., Klawans, H. L. & Koetsier, J. C.) Vol. 3 Ch. 47 213-257 (Amsterdam/New York, Elsevier Science, 1985).
6. Zamvil, S. S. & Steinman, L. Annu. Rev. Immunol. 8, 579-621 (1990). | PubMed | ISI |
7. Ota, K. et al. Nature 346, 183-187 (1990). | PubMed | ISI |
8. Lucchinetti, C. et al. Ann. Neurol. 47, 707-717 (2000). | PubMed | ISI |
9. McFarland, H. F. et al. Ann. Neurol. 32, 758-766 (1992). | PubMed | ISI |
10. Becker, K. G. et al. Proc. Natl Acad. Sci. USA 95, 9979-9984 (1998). | Article | PubMed | ISI |
11. The IFNB Multiple Sclerosis Study Group and the University of British Columbia MS/MRI Analysis Group. Neurology 45, 1277-1285 (1995). | PubMed |
12. Johnson, K. P. et al. Neurology 45, 1268-1276 (1995). | PubMed | ISI |
13. Duda, P. W., Schmied, M. C., Cook, S. L., Krieger, J. I. & Hafler, D. A. J. Clin. Invest. 105, 967-976 (2000). | PubMed | ISI |
14. Wandinger, K. P. et al. Ann. Neurol. 10 (in the press, 2001).
15. The Lenercept Multiple Sclerosis Study Group and the University of British Columbia MS/MRI Analysis Group. Neurology 53, 457-465 (1999). | PubMed |
16. Moalem, G. et al. Nature Med. 5, 49-55 (1999). | Article | PubMed | ISI |
17. Owens, T., Wekerle, H. & Antel, J. Nature Med. 7, 161-166 (2001). | Article | PubMed | ISI |
18. Hohlfeld, R. Brain 120, 865-916 (1997). | Article | PubMed | ISI |
19. Kennedy, M. K., Tan, L.-J., Dal Canto, M. C. & Miller, S. D. J. Immunol. 145, 117-126 (1990). | PubMed | ISI |
20. Nicholson, L. B., Greer, J. M., Sobel, R. A., Lees, M. B. & Kuchroo, V. K. Immunity 3, 397-405 (1995). | PubMed | ISI |
21. Bielekova, B. et al. Nature Med. 6, 1167-1175 (2000). | Article | PubMed | ISI |
22. Kappos, L. et al. Nature Med. 6, 1176-1182 (2000). | Article | PubMed | ISI |
23. Jacobs, L. D. et al. New Engl. J. Med. 343, 898-904 (2000). | PubMed | ISI |

© 2001 Nature Publishing Group