Multiple sclerosis (MS) is a heterogeneous disease clinically, pathologically and radiographically1, 2, 3, 4. Most patients (
85%) exhibit an initial relapsing–remitting (RRMS) course, and the majority will go on to develop secondary progressive MS (SPMS), which is characterized by worsening neurological disability with or without superimposed attacks. Primary progressive MS (PPMS) accounts for approximately 10% of patients, and involves continuous disease progression from onset, without relapse or remission5. Progressive relapsing MS is a rare subtype that is characterized by neurological progression from onset, with superimposed relapses thereafter.
The positive impact of disease-modifying therapies on clinical and imaging parameters of relapsing MS activity is now well established. Nevertheless, while active inflammation may determine the initial disease course, natural-history studies suggest that disease progression takes place independently of acute inflammation6. In particular, the hypothesis that dysregulation of the innate immune system may be a key factor involved in the shift towards irreversible neurodegeneration and MS progression is a provocative idea that may in part explain why currently approved MS therapies with anti-inflammatory mechanisms that target the adaptive immune system show little effect in progressive stages7 (Fig. 1). A leading hypothesis is that long-term disability may be due to irreversible neurodegeneration that begins early, even before progression may begin clinically, and that progression does not ensue until a critical threshold of axons has been destroyed. Although axonal damage is more prominent in active inflammatory lesions, independent axonal and neuronal loss can be observed early on in MS, not only within active focal lesions, but also in chronic silent plaques and in normal-appearing white and grey matter, as demonstrated pathologically8, 9, 10, 11, 12, 13, 14. Interestingly, neuronal loss and secondary cortical atrophy is more prominent, even early on, in progressive forms of MS, and neuroaxonal damage in PPMS seems to be partially independent from T2 lesion load15. So, insidious damage in normal-appearing parenchyma of the central nervous system (CNS) might yet be an additional factor that is implicated in MS progression. The specific biological mechanisms that underlie progression in MS are not fully understood, but the roles of immunological, biochemical, electrophysiological and genetic factors are under active investigation. Overall, the view has now evolved to MS being a constitutive diffuse syndrome, rather than the classical notion of a multifocal disease with periodic heightened immune activation.
Figure 1 | The immune response dichotomy in multiple sclerosis.
a | A key step in early disease — entry of self-reactive lymphocytes (adaptive immune effectors) into the central nervous system (CNS) — initiates an abnormal immunopathogenic cascade, which leads to relapsing focal damage. b,c | Disease progression may relate to an immune balance shift that is characterized by abnormal activation of innate immune effectors — such as dendritic cells, resident CNS microglia, macrophages and natural killer cells — and astrocytes together with upregulated production of pro-inflammatory cytokines and toxic molecules — tumour necrosis factor-
(TNF-), interferon-
(IFN-), T-cell immunoglobulin mucin 3 (TIM3; also known as HAVCR2), galectin 9 (GAL9; also known as LGALS9), glutamate, and reactive oxygen species (ROS) and reactive nitrogenous species (RNS). c | An additional element that is implicated in the progression of multiple sclerosis may be the presence of ectopic meningeal B-cell follicles. Abnormally activated B cells present in these structures produce chemokines that further attract other immune effectors. The events described in (b,c) and (d), either in conjunction or independently, may perpetuate diffuse CNS inflammation and neurodegeneration. CCL2, chemokine (C-C motif) ligand 2; CXCL13, chemokine (C-X-C motif) ligand 13; iNOS, inducible nitric oxide synthase; IL10, interleukin 10; RRMS, relapsing–remitting multiple sclerosis; SPMS, secondary progressive multiple sclerosis; TGF-
, transforming growth factor-; TH1 cell, T helper 1 cell; TReg cell, regulatory T cell.
MS is classically described as a chronic inflammatory disease of the CNS marked by focal autoreactive T-cell and macrophage infiltrates that lead to demyelination, and axonal and neuronal loss16. The accumulation of activated microglia and macrophages is a feature in the later development of lesions, whereas blood–brain barrier damage, prominent infiltration by activated CD4+ T cells and clonotypic CD8+ T cells, and the presence of reactive astrocytes and proliferating oligodendrocytes are characteristic of acute plaques17, 18, 19. The presence of these cells, in addition to pro-inflammatory cytokines such as interleukin 12 (IL12) and tumour-necrosis factor- (TNF-)20, 21, 22, 23, has led to the classical hypothesis that MS is a T helper 1 (TH1)-cell-mediated autoimmune disease. However, this concept is now challenged by mounting evidence that other adaptive immune cells (such as TH17 cells and peripheral B lymphocytes), as well as key innate immune cells (dendritic cells, natural killer T cells and resident microglia), also play a significant role in MS pathogenesis. These could represent new potential targets for MS therapeutics (Fig. 2).
Figure 2 | Multiple sclerosis immunopathogenesis and therapeutic targets.
Immature dendritic cells are central players in innate immune responses and are involved in the maintenance of peripheral tolerance by means of promotion of suppressor regulatory T cell and anti-inflammatory T helper 2 (TH2) cell responses. Abnormally activated (mature) antigen-presenting dendritic cells can be found in patients with multiple sclerosis (MS). This activation results in increased production of pro-inflammatory cytokines that lead to aberrant activation of TH1 and TH17 pro-inflammatory responses. Activated encephalitogenic adaptive immune effectors (such as TH1 cells, TH17 cells, CD8+ cells and B cells) express surface molecules that allow them to penetrate the blood–brain barrier and to enter the central nervous system (CNS). The presence of autoreactive immune effectors, together with abnormally activated CNS astrocytes and microglia, lead to increased production of reactive species, excitotoxicity, autoantibody production and direct cytotoxicity, which are all involved in demyelination, axonal and neuronal damage that is present in patients with MS. Potential therapeutic interventions at different levels of the immunopathological cascade are depicted in filled yellow boxes. CTL, cytotoxic T lymphocyte; IFN-, interferon-; IL, interleukin; IL2R, IL2 receptor; MHC II, major histocompatibility complex class II; MMPs, matrix metalloproteinases; NO, nitric oxide; ODG, oligodendrocyte; S1PR, sphingosine 1-phosphate receptor; TGF-, transforming growth factor-; TNF, tumour-necrosis factor; TReg cell, regulatory T cell; VCAM1, vascular cellular adhesion molecule 1; VLA4, very late antigen 4 (also known as 41 integrin).
Clinicians treating MS therefore encounter a multifaceted adversary, not only based on its variable clinical presentation, but also regarding striking clinical and radiological dissociation, and different pathological subtypes (Box 1). This heterogeneity contributes to the presently limited ability to predict functional outcome in a significant number of patients — who are often treated on a 'hopeful empirical' basis, especially in progressive stages — for which US Food and Drug Administration (FDA)-approved disease-modifying therapies (which are briefly discussed in Box 2) have not demonstrated a clear benefit.
In this Review, we discuss novel MS therapeutic agents based on their targeting of specific steps in the MS immunopathogenic cascade, including different aspects of T-cell and B-cell function, mixed targets, and neuroprotective strategies.
Targeting leukocyte trafficking
Clinical relapses and new brain magnetic resonance imaging (MRI) lesions in MS are associated with the active infiltration of autoreactive lymphocytes and monocytes into the CNS. A key step in gaining access to the CNS is leukocyte migration and disruption of the blood–brain barrier. Leukocyte recognition of specific target antigen/human leukocyte antigen (HLA) class I complexes expressed by CNS cells activates the inflammatory cascade, which leads to CNS damage. So, disruption of leukocyte trafficking and migration across the blood–brain barrier could potentially have a beneficial effect on clinical and radiological disease activity.
ADVERTISEMENT
Natalizumab. Early studies in experimental autoimmune encephalomyelitis (EAE) animal models suggested that blockade of surface adhesion molecules curtails T-lymphocyte trafficking across the blood–brain barrier and suppresses clinical disease24, 25, 26, 27. Natalizumab (Tysabri; Elan/Biogen Idec) is a humanized monoclonal antibody (mAb) that is directed against the 4 subunit of the integrin VLA4 (very late antigen 4) adhesion complex, which is normally expressed on activated lymphocytes, monocytes and other cell types. As patients with MS show a threefold to fourfold increase in VLA4 expression in cerebrospinal fluid (CSF) and in circulating lymphocytes compared with normal subjects, this complex appears to be pathologically relevant in this disease28. Although the specific in vivo immune effects of natalizumab remain unclear, it is thought that impaired migration of T cells, B cells and monocytes into the CNS results from blockade of VLA4 binding to vascular cell-adhesion molecule 1 (VCAM1) expressed on the brain microvasculature29. In addition, VLA4 seems to be involved in T-cell co-stimulation, and anti-VLA4 signals can modulate T-cell activation both in EAE animal models and in humans. Furthermore, natalizumab may also modulate immune-cell proliferation and disrupt T-cell activation by blocking the interaction of VLA4 with the extracellular matrix30, 31, 32, 33.
One Phase II and two Phase III clinical trials in patients with MS have demonstrated the beneficial effects of natalizumab in acute MS relapses. Although treatment with this drug did not affect the actual clinical course of relapse, a two-thirds decrease in relapse rate and significant reduction in disability progression and MRI lesion load were demonstrated34, 35. Based on these results, the FDA approved natalizumab in November 2004 for the treatment of RRMS. Although this drug was well tolerated by patients in general, the unfortunate development of three cases of progressive multifocal leukoencephalopathy (PML)36, 37 (two fatal) — all in the context of concomitant immunomodulatory treatment — led to the suspension of clinical trials in February 2005. It was estimated that the incidence of PML among all patients treated with natalizumab in trials for different autoimmune diseases was 1 in 1,000 (Ref. 38). Although the pathogenesis of PML in these patients remains unclear, it is known that VLA4 is expressed at high levels in B cells and in CD8+ lymphocytes, and B cells have been implicated as the main systemic carriers and transporters of JC virus across the blood–brain barrier. PML development in natalizumab-treated patients may relate to anti-VLA4-mediated disruption of the migratory and antiviral properties in these specific immune-cell subsets. After re-evaluation of 2 years of safety and efficacy data by the FDA, natalizumab was re-introduced into the market in 2006 as a monotherapy for patients with relapsing forms of MS who have not responded to or cannot tolerate other treatments for MS. In August 2008, two new PML cases in patients receiving natalizumab monotherapy for over a year were reported in Europe. Thus, careful patient monitoring should be followed while using this drug.
FTY720 (fingolimod). A synthetic analogue of the fungal sphingosine 1-phosphate (S1P) receptor agonist myriocin, FTY720, is a potential novel oral immunosuppressant for MS.
Under normal circumstances, autoreactive naive T cells undergo negative selection that leads to tolerance, but in patients with autoimmune diseases such as MS, these cells persist. When specific naive T cells are presented with CNS self-antigens by dendritic cells in the lymph nodes, the naive T cells become activated encephalitogenic effector cells that have the potential to initiate an attack upon egress into the CNS. FTY720 induces lymphocyte homing and sequestration into the lymph nodes39, which precludes the systemic trafficking of self-reactive T cells and their CNS invasion. This effect is mainly achieved through the interaction of FTY720 with G-protein-coupled S1P receptors. In particular, the S1P1 receptor is predominantly expressed on lymphocytes and regulates their migration from secondary lymphoid tissues. Upon phosphorylation and S1P receptor binding, FTY720 initially acts as an agonist, but it then induces downregulation of receptor expression, therefore blocking the signal required for lymphocyte migration.
Unlike other traditional systemic immunosuppressants, FTY720 does not seem to adversely affect T-cell function in response to viral infections. Interestingly, experimental evidence suggests that FTY720 differentially affects the sequestration of CD4+CD25+ regulatory T (TReg) cells and upregulates their suppressive function40, 41. Furthermore, in addition to its effect on lymphocyte homing, this drug may directly regulate dendritic cell maturation42, 43, and specifically downregulate pro-inflammatory signals41. The extent to which FTY720 affects S1P-mediated regulation of oligodendrocyte maturation, myelinogenesis and astroglia proliferation, and their relationship to its therapeutic effectiveness and safety, remains to be elucidated44, 45.
In a recent proof-of-concept study, daily use of oral FTY720 in patients with RRMS showed a statistically significant benefit on time to first relapse, annual relapse rate, total T2 lesion load and new number of new enhancing lesions on MRI compared with placebo. No impact on the change in expanded disability status scale (EDSS) over time was demonstrated46. Two Phase III trials are currently underway to further evaluate this promising new agent versus placebo (FREEDOMS study) or interferon-1a (IFN-1a) (TRANSFORMS study) in patients with RRMS. The role of this drug in PPMS will also be evaluated in a Phase III trial that is comparing high-dose FTY720 versus placebo.
Targeting T cells
Active demyelinating MS lesions are characterized by preferential perivascular and meningeal accumulation of autoreactive CD4+ T cells (B cells and plasma cells are also present but in smaller numbers), whereas CD8+, 
T cell and V2 T cells, and macrophages invade the lesional parenchyma. T-cell receptor-mediated recognition of self-CNS antigens in the context of major histocompatibility complex (MHC) class II presentation results in CD4+ T-cell clonal expansion and activation that is skewed towards a TH1 pro-inflammatory response.
Accumulating evidence suggests a role for cytotoxic CD8+ T cells in the pathogenesis of MS. CD8+ T-cell-induced CNS damage is dependent on MHC class I recognition of neuronal and oligodendrocyte self-antigens or, potentially, certain molecular mimickers. CD8+ T cells are more abundant in MS brain parenchyma than CD4+ T cells, and clonotypic expansions of memory CD8+ T cells are present in the CSF, brain tissue and blood of patients with MS47, 48. Furthermore, CD8+ T lymphocytes display increased reactivity to myelin antigens, and secrete chemokines for myelin-specific CD4+ T cells in MS49, 50. An increase in lymphotoxin production by CD8+ T cells in SPMS51 and a correlation between cytokine CD8+ cytokine production and CNS damage on MRI scans have also been reported52.
Daclizumab. IL2 is the main growth factor for activated T lymphocytes, which stimulates their clonal expansion and maturation. Daclizumab (Zenapax; Roche) is a humanized mAb against the IL2 receptor- (IL2R) chain (CD25). Although this cytokine receptor is rarely expressed in normal T cells, activated T cells do so at high levels. Daclizumab received initial approval by the FDA for renal-transplant rejection, but it has been safely used off-label in organ transplantation, selected T-cell-mediated autoimmune diseases and haematological malignancies, as elevated soluble IL2R levels can be detected in a range of such disorders53. IL2R blockade results in the inhibition of activated T-cell proliferation and expansion in vitro, whereas inhibition of activated T lymphocytes expressing IL2R in rodent EAE models results in clinical improvement54, 55.
With the intent of specifically targeting self-reactive T cells, daclizumab has been tested in MS. In an open-label baseline versus treatment Phase II trial involving IFN- non-responder patients with RRMS, and SPMS patients with highly active disease, treated subjects had a pronounced (78%) reduction in brain inflammatory activity compared with baseline, together with stabilization of clinical-disease progression56. In contrast to in vitro data, further analysis of these patients suggests that daclizumab's CNS anti-inflammatory effect correlates with the striking IL2-mediated specific activation and expansion of CD56bright natural killer cells — innate immune effectors that presumably negatively regulate CD4+ and CD8+ T-cell survival in the treated patients57.
CHOICE, the first Phase II, randomized, double blind, placebo-controlled clinical trial, enrolled 230 patients with active, relapsing MS58. Treatment groups received subcutaneous daclizumab at 1 mg kg-1 every 4 weeks or 2 mg kg-1 every 2 weeks for a total of 24 weeks, in addition to concurrent IFN- therapy. Preliminary analysis on week 24 data showed that the study's primary efficacy end point — total number of new or enlarged gadolinium-positive lesions — was significantly reduced by 72% (p = 0.004) in the high-dose daclizumab group compared with placebo. The low-dose treatment group had a 25% reduction that was not statistically significant. Overall, daclizumab was considered to be safe and well tolerated, although a higher incidence of serious infections and mild-to-moderate cutaneous adverse events were seen in those receiving daclizumab. The SELECT trial, a Phase II trial comparing low-dose and high-dose stand-alone daclizumab therapy with placebo in patients with active, relapsing MS, is planned.
Targeting TReg cells. Thymic and peripheral induction of FOXP3+CD4+CD25+/- TReg cells, balanced with that of effector cells, is essential for the maintenance of immune tolerance. TReg cells exert their tolerogenic effect via suppression of effector T-cell activation and proliferation. Disruptions in the function of TReg cells result in a shift of immune homeostasis towards inflammation and autoimmunity.
Patients with RRMS display a loss of functional suppression by TReg cells in vitro that correlates with decreased FOXP3 expression59, 60, 61, 62, and similar abnormalities have been described in other autoimmune diseases. This defect in TReg cell function is not associated with MS relapses60 and, interestingly, such impairment has not been demonstrated in patients with SPMS62.
TReg cells seem to be an attractive target for tolerogenic MS immunotherapy, and indeed strategies aimed at their ex vivo and/or in vivo functional activation and expansion are under active investigation. Ideally, to re-establish the regulatory versus autoreactive effector T-cell imbalance involved in MS pathogenesis, the synergistic combination of agents that induce the former together with compounds that suppress the latter is a logical approach. Along this line, rapamycin (sirolimus), a drug widely used for preventing organ transplant rejection, enriches for functional FOXP3+CD4+CD25+ TReg cells, whereas CD4+ effector T cells are depleted by this agent63, 64, 65, 66, 67, 68. This effect on the TReg cell/effector cell ratio may therefore be exploited in MS therapy.
Some established MS drugs positively affect TReg cell function in MS. For example, the unique mode of action of glatiramer acetate (Copaxone; Teva Pharmaceutical) is in part mediated by the induction of FOXP3+CD4+CD25+ TReg cell proliferation and migration across the blood–brain barrier69. Other immunoregulatory effects of glatiramer acetate include the sustained peripheral induction of a specific TH2 shift in response70, 71, 72 and induction of CD8+ suppressor T cells73, 74. Effects also include in situ CNS bystander suppression of inflammation and release of neurotropins75, 76, 77. As mentioned above, FTY720 also enhances CD4+CD25+ TReg cell function40, 41. Together with other experimental lines, antigen-specific oral tolerance induction has been shown to ameliorate and/or prevent autoimmune disease in animal models of diabetes78, 79, encephalomyelitis80, atherosclerosis81, lupus82, 83 and colitis84 in which transforming growth factor- (TGF-)-dependent induction of TReg cells seems to have a central role in disease inhibition.
The gender predilection observed in MS and other autoimmune diseases85, along with the protective effects of pregnancy on disease exacerbations86, implicate sex hormones in autoimmune pathogenesis. In particular, treatment with the pregnancy hormone oestrogen (E2) has been shown to suppress EAE87, 88, 89. Interestingly, in this model, oestrogen not only directly inhibits self-reactive antigen-presenting cells, effector T cells and their transmigration across the blood–brain barrier, but it also enhances FOXP3+ TReg cell function90, 91, 92. Preliminary studies also hint at a positive effect of oestrogen treatment in patients with MS93. Further clinical investigations on its potential as a therapeutic adjuvant in MS are underway.
Antigen-specific therapies. Binding of the antigen-specific T-cell receptor (TCR) to peptide/MHC class II complexes on the surface of antigen-presenting cells triggers the initial activation of T cells. Additional co-stimulatory signals are required for full T-cell activation, while TCR crosslinking alone results in partial activation that promotes CD4+ T-cell apoptosis, anergy or deletion, which leads to antigen-specific tolerance. Based on this premise, therapies targeted at inactivation of autoreactive T cells and restoration of self-tolerance have been developed. Although the highly specific nature of these therapies offers the advantage of bypassing undesired global immunosuppressive effects, their potential for inducing specific allergic reactions or worsening autoimmunity poses safety concerns. In addition, as pathogenic T cells in patients with MS display a heterogeneous self-antigen reactivity repertoire, targeting a single specific antigen may result in limited efficacy. We will discuss below several experimental approaches that are ultimately aimed at maximizing therapeutic effectiveness and safety in patients with MS. These include peptide-specific tolerance; anti-CD3 mAbs; and DNA, T cell and TCR vaccination and gene-delivery strategies.
With regard to peptide-specific tolerance, transmucosal administration of tolerogens offers advantages in terms of ease of delivery and limited toxicity. Oral myelin basic protein (MBP) effectively induces dose-dependent tolerance in EAE models94. A double-blind pilot MS trial of oral tolerization with bovine myelin antigens showed a reduction in number of myelin-reactive T cells95; however, a subsequent Phase III trial with oral bovine myelin did not show clinical benefit. Nasal tolerization restores self-tolerance in EAE models, but no MS clinical trials have taken place to date96, 97, 98, 99. Although mucosal delivery of soluble myelin peptides protects against EAE induction, when attempted after clinical disease onset it may have less efficacy98, 99, 100, 101, 102, 103. With the intent of inducing clonal deletion of myelin-specific self-reactive T cells, intravenous administration of either native myelin peptides or soluble HLA–peptide complexes has been explored, and these have demonstrated tolerogenic efficacy in animal models of MS104, 105, 106, 107, 108.
MBP8298 is a synthetic myelin basic peptide that is currently under investigation in MS with the intent of inducing high-dose tolerization to MBP, one of the main potential immune targets of self-reactive B cells and T cells in MS. In a 24-month placebo-controlled, double-blinded Phase II trial, intravenous MBP8298 (500 mg) or inactive placebo was administered twice a year to 32 patients with SPMS or PPMS109. Although no significant clinical difference was observed between the treatment and placebo groups, a subgroup of HLA-DR2 and/or DR4-positive patients did benefit from treatment, as reflected by a statistically significant delay in clinical progression (as measured by changes in EDSS). Specifically, the median time to progression was 78 months for MBP8298-treated patients compared with 18 months for the placebo group. Interestingly, both HLA-DR2 and DR4 haplotypes have been linked to increased susceptibility to MS. It should be noted that although MBP8298 treatment induced the suppression of MBP autoantibodies in CSF, this biological effect did not correlate with clinical stabilization. With regards to adverse reactions, MBP8298 demonstrated a safe profile. Further evaluation of MBP8298 in MS is currently underway in the Phase III clinical trials MAESTRO-03 and MAESTRO-01 for patients with SPMS and in the European Phase II trial MINDSET-01 for RRMS.
The use of syngeneic fixed splenocytes that are coupled with encephalitogenic peptides is an alternative approach to circumvent the efficacy and safety limitations posed by oral and soluble peptide tolerogens110. This strategy is thought to induce T-cell anergy, activation of regulatory T cells, and bystander suppression, with all of them contributing to the re-establishment of self-tolerance. Treatment in different animal models of autoimmune disease, including EAE, prevents both clinical disease onset and pre-existing disease progression111, 112, 113, 114, 115. A Phase II study in patients with new onset RRMS is currently awaiting FDA approval.
Aberrant T-cell responses against native CNS peptide antigens are implicated in the pathogenesis of MS. Tolerance to these CNS peptide antigens can be induced with altered peptide ligands (APLs), which work by inducing encephalitogenic T-cell anergy, inducing TH2 cell phenotype shifts and bystander suppression, among other mechanisms116, 117, 118, 119, 120, 121. APLs can prevent the development of EAE in mice, but their therapeutic potential in MS remains uncertain. Two initial Phase II trials with an MBP83–99-derived APL raised safety considerations owing to unexpected disease exacerbation122 and hypersensitivity reactions123, although some preliminary clinical and MRI disease activity measures were positive. As MS is a disease with abnormal immune responses against not just one but a heterogeneous group of CNS peptide antigens, the use of APL mixtures, rather than a single peptide, may provide a broader and more biologically relevant effect.
TCR crosslinking in the absence of co-stimulation leads to T-cell anergy. A human-specific anti-CD3 mAb, muromonab-CD3 (Orthoclone OKT3; Ortho Biotech), has been widely used in the treatment of organ transplantation rejection and in certain autoimmune diseases. As a beneficial autoreactive T-cell-depleting effect was anticipated, muromonab-CD3 was tested in an MS clinical trial124. Treated patients suffered disease exacerbation that was correlated with nonspecific induction of increased TNF- and other pro-inflammatory cytokines125. The use of non-mitogenic anti-CD3 mAbs in EAE suppressed disease, with less T-cell activation and cytokine production126, 127. We have found that oral administration of anti-CD3 mAb both prevents and treats ongoing EAE by inducing a TGF-b dependent Treg cell128. There was no cytokine release or T cell activation. Oral anti-CD3 is also effective in models of diabetes129 and lupus130. A Phase I trial of oral-CD3 in healthy volunteers showed no toxicity. Immunologic effects including a decrease of TH17 responses and an increase of TGF-b were observed (H. Weiner and Y. Ilan, unpublished observations). Trials of oral anti-CD3 are planned in MS. Intravenous non-mitogenic anti-CD3 has shown positive results in Type I diabetes131, 132.
Myelin DNA cocktail vaccination and myelin antigen gene therapy are exquisitely specific tolerogenic therapies that can induce clinical and pathological disease protection in acute EAE models133, 134, 135, 136, 137, 138. Myelin DNA vaccination in chronic EAE mouse models was initially shown to be effective in reducing the relapse rate, but not on disease progression or long-term disability139, 140. In this model, chronic DNA vaccination resulted in TH1-cell-biased immunostimulation, which is probably secondary to Toll-like receptor stimulation141 via immunogenic CpG sequences in the DNA vaccine construct backbone. This undesired effect could be counteracted by co-vaccination with TH2 cytokine-coding constructs and/or synthetic antagonistic oligodeoxynucleotides, which act synergistically to induce antimyelin TH2-cell proliferation. Reduced relapse rate and mean disease severity in treated animals were demonstrated with this strategy140, 142.
Several groups have demonstrated the tolerogenic potential of autologous vaccination with attenuated or inactivated encephalitogenic CD4+ or CD8+ T-cell clones in EAE models143, 144. Preliminary MS vaccination trials with attenuated autologous peripheral or CSF myelin-specific T cells suggested that this strategy induces clonotypic depletion of MBP-reactive T cells. Some studies have shown decreased relapse rate and MRI lesion load, but no significant impact on time to progression to disability145, 146, 147, 148. Protection against EAE can also be achieved by vaccination with synthetic TCR peptides similar to those that are specifically expressed by encephalitogenic T cells149, 150. Pilot studies have demonstrated their safety and immunogenicity in patients with MS151, 152, 153, 154, 155, and may translate into future therapeutic testing.
Targeting B cells
The role of the humoral immune response in MS pathogenesis is supported by several observations. Although CNS demyelination has only been reproduced in animal models by means of adoptive transfer of autoreactive T cells, EAE studies have shown that B cells are involved in disease initiation156 and can also modify disease phenotype and severity157.
In patients with MS, high numbers of chronically activated B cells, plasma cells and increased immunoglobulin M (IgM) and IgG antimyelin antibodies are present in the CSF and meninges and seem to correlate with the rate of disease progression158, 159, 160, 161. Furthermore, B-cell clonal expansion and oligoclonal IgG production can be detected in brain plaques and CSF from patients with MS162, 163, 164, 165, 166, 167. Myelin-oligodendrocyte-glycoprotein-specific antibodies have also been isolated from MS lesions168. Reported overexpression of the survival factor BAFF (also known as TNFSF13B) in the brains of patients with MS may account in part for these observations169; that is, BAFF potentially mediates prolonged survival of autoreactive B cells in the CNS. Additionally, antibody deposition and complement activation characterize MS lesions with a type II pathological pattern170, whereas plasmapheresis may show benefit in patients with this pathological phenotype171.
B cells may have a role in MS pathogenesis not only through production of self-reactive antibodies, but also through their antigen-presenting and cytokine-secreting role, which can lead to abnormal T-cell and macrophage activation, thereby perpetuating CNS inflammation and damage. Depletion of these autoreactive B cells is therefore an attractive therapeutic strategy in MS.
Rituximab (Rituxan/MabThera; Biogen Idec/Genentech/Roche) is a chimeric murine–human IgG1
mAb against CD20+ pre-B cells and mature B cells. Memory B cells are key mediators of long-term humoral immunity. In patients with MS, high numbers of autoreactive memory B cells can be found intrathecally and may participate in the perpetuation of CNS damage in progressive MS stages. Memory B cells express CD20, but do so at a lower level than naive B cells. Although it remains unclear to what extent rituximab affects different immune-cell populations and how it correlates with its potential efficacy in MS, recent studies suggest that treatment with rituximab may lead to reversible depletion of circulating B cells by induction of apoptosis, antibody-dependent cellular-mediated cytotoxicity and/or complement-induced cytolysis (see Fig. 3 for a summary).
Figure 3 | Rituximab: potential mechanisms of action in multiple sclerosis.
Binding of the monoclonal antibody rituximab to CD20 on the surface of B cells may result in immune modulation towards anti-inflammation owing to potential inhibition of B-cell activation and functions (that is, antigen presentation, effector T-cell activation, plasma-cell differentiation and autoantibody production) as well as the induction of activated regulatory T (TReg) cells. B cell numbers are decreased via induction of apoptosis. IgG, immunoglobulin G; IL6, interleukin 6; TH, T helper; T regulatory type 1 (TR1).
The implication of B cells in MS pathogenesis goes beyond the simplistic view of these cells as mere autoantibody producers. It is now well established that B cells also act as efficient antigen-presenting cells and cytokine producers, and are involved in key steps of effector T-cell activation and induction of regulatory T cells172, 173, 174, 175. B-cell-depleting agents should be expected to alter all aspects of B-cell participation in the immune response, and T and B cell cross-talk in particular, therefore potentially broadening their therapeutic effects in MS.
Rituximab is currently approved by the FDA for the treatment of refractory B-cell non-Hodgkin's lymphoma and rheumatoid arthritis, but it has also been used off-label in other haematological, rheumatic and autoimmune diseases. Rituximab has also demonstrated clinical benefit in anti-GM1 (monosialotetrahexosylganglioside) polyneuropathy176. Several small non-blinded studies have reported beneficial effects in some patients with neuromyelitis optica, fulminant RRMS and PPMS177, 178, 179. In a recent Phase II, randomized, double-blind, placebo-controlled 48-week trial, patients with RRMS received two rituximab intravenous infusions (1,000 mg each) — one on day 1 and one on day 15 of the study period180. A statistically significant 91% reduction in the primary end point — total number of gadolinium-enhancing T1 lesions on brain MRI scans at weeks 12, 16, 20 and 24 — was observed in the rituximab-treated versus placebo group (p<0.001). The proportion of patients with relapses during 24 weeks was also significantly reduced in the rituximab-treated group (p = 0.02)180. With regards to efficacy in progressive stages, a recent Phase II/III, randomized, double-blind, parallel-group, placebo-controlled study in 439 patients with PPMS treated with four courses of rituximab or placebo 6 months apart failed to significantly delay the time to confirmed disease progression. Additional large-scale controlled studies are required to define which particular immunophenotypical and clinical-stage patient subsets may benefit from rituximab use. A Phase III trial is currently ongoing.
Although rituximab is a relatively safe drug, its manufacturer recently reported the incidence of two fatal PML cases in patients treated with multiple courses of rituximab for severe haematological manifestations of systemic lupus erythematosus. Both patients were on chronic steroid immunosuppression, and had received azathioprine and cyclophosphamide previously. PML has been previously reported in 23 patients with haematological malignancies after receiving rituximab — almost all in combination with chemotherapy and/or stem-cell transplantation. So far, a clear link between PML and independent rituximab administration has not been established. The consensus on this issue seems to be that PML was attributable to severe immunosuppression due to combined therapy, rather than to rituximab alone. Nevertheless, care and monitoring should be maintained in patients treated with this agent.
Mixed targets
Alemtuzumab. Alemtuzumab (Campath; Genzyme/Bayer) is a humanized mAb against CD52, a protein that is present on the surface of T cells, B cells and monocytes. Alemtuzumab induces the immediate cytolysis of these cells, thereby depleting the number of autoreactive lymphocytes with potential to reach the CNS. This marked lymphopaenic effect is prolonged but reversible over time. In a small observational clinical study, alemtuzumab demonstrated a significant effect in reducing active CNS inflammation, and annual relapse was rare in patients with RRMS and those with SPMS. However, while patients with RRMS treated with this agent had a modest delay in disease progression, this effect was not observed in patients with SPMS181, 182. This difference in response between disease stage is not entirely unexpected, and is consistent with the current thought that CNS inflammation is independent from axonal loss and neurodegeneration, with the latter being crucial factors implicated in disease progression and disability.
In the CAMMS223 trial — an open-label, rater-blinded Phase II study — 334 treatment-naive patients with early active RRMS were randomized to receive up to three annual cycles of either low-dose alemtuzumab (12 mg per day) or high-dose alemtuzumab (24 mg per day) versus high-dose subcutaneous IFN-1a, in addition to intravenous methylprednisolone for all patients183. Interim 2-year analysis revealed a 72% and 88% decrease in the cumulative number of relapses (both p< 0.0001), as well as 66% and 88% delay in disability (multiple sclerosis functional composite score; MSFC), in the groups receiving low-dose or high-dose alemtuzumab (both p<0.01), respectively, compared with IFN- treatment184. A Phase III trial comparing subcutaneous IFN-1a with alemtuzumab is scheduled.
Results from an additional open-label, investigator-initiated Phase II study in patients with active RRMS refractory to IFN- receiving two annual cycles of alemtuzumab (192 mg total dose) further support its efficacy. A 94% reduction in relapse rate (p<0.0001) and stabilization or improvement in disability (EDSS) and functional status (MSFC) in 70% to 87.5% of the patients, demonstrate the benefits of alemutuzmab185.
Alemtuzumab is associated with an increased risk of developing immune thrombocytopenic purpura and thyroid disorders.
Laquinimod. Laquinimod (ABR-215062) is a novel oral synthetic immunomodulator that has demonstrated efficacy in acute and chronic EAE186. Recently, laquinimod treatment has also been shown to inhibit the development of chronic EAE in an Ifn -knockout mouse model187. The specific mode of action behind this drug's anti-inflammatory effect remains to be elucidated, but treated animals reportedly display decreased CD4+ lymphocyte and macrophage CNS infiltration, and a shift towards a TH2/TH3 cell immunophenotype188. In humans, a proof-of-concept study in patients with RRMS receiving laquinimod daily for 2 years showed no difference in clinical relapse rate and disability, but there was a significant dose-dependent reduction in active lesion load on MRI compared with controls189. This drug is safe and well tolerated, without general immunosuppressive effects. Based on initial data, laquinimod shows reasonable promise as a novel oral MS therapy, either as a synergistic add-on to IFN or as an alternative agent in IFN- non-responders.
Cladribine. Cladribine (2-chlorodeoxyadenosine; 2-CdA) is an adenosine deaminase-resistant nucleoside analogue that has selective lymphotoxicity. Two independent Phase II, placebo-controlled clinical trials have shown stabilization of disability and a reduction in MRI measures of disease activity in cladribine-treated patients with RRMS or SPMS, but not those with PPMS190, 191, 192. CLARITY, the first pivotal Phase III study of oral cladribine as monotherapy in patients with RRMS, and ONWARD, a similar Phase III study of cladribine as add-on therapy to new formulation of IFN- in patients with active MS, are currently underway.
BG-12. BG-12 is a second-generation oral fumarate with reported immunomodulatory and neuroprotective actions193. In a Phase II double-blind, placebo-controlled trial, 257 patients with RRMS were randomized to receive 120 mg per day, 360 mg per day or 720 mg per day of BG-12 or placebo for 6 months194. Patients receiving 720 mg per day had a statistically significant decrease (69%) in the total number of gadolinium-positive brain MRI lesions and a reduction in the number of new or enlarging T2 and hypointense T1 lesions. A trend towards reduction (32%) in relapse rate was also seen in the group receiving 720 mg per day. Overall, BG-12 was safe and well tolerated; no immunosuppression was detected in treated subjects at all doses.
Two ongoing Phase III studies in patients with RRMS, CONFIRM and DEFINE, are comparing BG-12 with placebo or glatiramer acetate, and BG-12 with placebo alone, respectively.
Teriflunomide. Teriflunomide, an inhibitor of pyrimidine synthesis with antiproliferative activity, is the active metabolite of leflunomide, a low-molecular-mass synthetic drug currently used to treat rheumatoid arthritis. Teriflunomide has shown a positive effect in the EAE model.
In a 36-week Phase II study in patients with RRMS or SPMS receiving low (7 mg) or high (14 mg) doses of teriflunomide or placebo, participants showed a significant decrease in the number of active lesions (new T2, enlarging T2 or gadolinium-positive T1 lesions) after 12 weeks of treatment195. Significantly fewer people in the group taking the higher dose of teriflunomide experienced an increase in EDSS when compared with the group given placebo. Currently, teriflunomide is being evaluated in an ongoing Phase III trial.
Mycophenolate mofetil. Mycophenolate mofetil is an antimetabolite prodrug that has immunosuppressive effects, which are mediated by multiple mechanisms including the cytostasis196, apoptosis and impairment of endothelial adhesion and transmigration of T cells and B cells197, among others. Mycophenolate mofetil may also act as a neuroprotectant by attenuating inducible nitric oxide synthase-mediated excitototoxicity. Its use in EAE accelerates clinical recovery198, and a Phase II trial in patients with RRMS suggests that mycophenolate mofetil plus IFN-1a in combination may reduce relapse rate and stabilize disease progression199, 200, 201.
Neuroprotection
Glutamate receptor modulation. Even at early stages, oligodendrocyte, neuronal and axonal degeneration are found in lesions and normal-appearing brain parenchyma in patients with MS. These neurodegenerative features seem to correlate with disease progression, independently from ongoing inflammation.
Increasing evidence implicates glutamate-mediated neurotoxicity in the pathogenesis of CNS diseases including amyotrophic lateral sclerosis, myelopathy, head injury, cerebral ischaemia and MS. The actual glutamate receptor type and level of expression on specific CNS cell types remains controversial, but evidence suggests that oligodendrocyte, astroglial, microglial and neuronal dysfunction, as seen, for example, in neuroinflammatory disease, affects glutamate metabolism, which then leads to subsequent excitotoxicity202, 203, 204. Furthermore, evidence suggests that glutamate may influence T-cell activation and transmigration, thus facilitating CNS cellular inflammation205.
The involvement of glutamate in EAE, and by inference in MS, is suggested by multiple reports of EAE amelioration following treatment with different glutamate receptor antagonists206, 207, 208, 209. It is also known that higher glutamate levels are present in the CSF of patients with active RRMS or with progressive SPMS, and seem to correlate with disease severity210, 211. By contrast, the expression of glutamate transporters, which are involved in the cellular uptake of excess glutamate, is reduced in the white matter of MS brains212. We have found that a fullerene attached to an N-methyl-D-aspartate (NMDA) receptor antagonist prevents axonal loss and demyelination in a model of progressive EAE by its effect on glutamate toxicity and nitric oxide212. Treatment reduced CCL2 expression on astrocytes and infiltration of CD11b cells into the brain.
Riluzole is a neuroprotective glutamate receptor modulator that inhibits glutamate synaptic release, thus presumably decreasing both its excitotoxicity and potentially inhibiting glutamate-mediated T-cell activation and transmigration. Prophylaxis with this agent resulted in clinical attenuation of EAE, with a correlating decrease in CNS inflammation, demyelination and axonal loss213. In humans, a small run-in riluzole trial in patients with PPMS showed a reduction in cervical-cord atrophy rate and new T1 black holes, whereas the rate of brain atrophy and T2 lesion load were not affected214, 215. This suggests that riluzole has an effect on axonal loss and lesion progression rather than on inflammation and new lesion development214, 215.
Minocycline is an oral semi-synthetic tetracycline antibiotic that can penetrate the CNS, and has interesting pleiotropic biological effects216. Its neuroprotective effect has been reported in animal models of neurological disease217, 218, 219, 220, 221, 222, 223 including CNS demyelination. Studies in EAE models have shown suppression of disease activity and disease progression both with prophylactic or therapeutic minocycline treatment224, 225, 226, 227. Potential mechanisms involved in its neuroprotective effect include reduced glutamate excitotoxicity by means of upregulation of glutamate transporters, regulation of intracellular signalling cascades, induction of an anti-apoptotic shift221, 228, and decrease in microglial activation225, 226, 229, 230. Additionally, minocycline has anti-inflammatory effects that are associated with T-cell proliferation attenuation, enhanced peripheral TH2 cell responses225, 230 and reduced lymphocyte CNS recruitment by inhibition of matrix metalloproteinases231. These effects correlate with reduction in pathological CNS inflammation and demyelination.
Combination therapy in EAE models with minocycline and either glatiramer acetate or IFN- results in synergistic attenuation of clinical disease and CNS pathology including inflammation, demyelination and axonal loss232, 233. When a 6-month minocycline regiment (100 mg orally twice daily) was given to a small group of patients with RRMS, it resulted in reduction of disease activity (enhancing lesion number) on MRI234. The ease of administration, low cost and good long-term use safety record235, 236 of minocycline, in addition to the promising preliminary evidence presented here, argue for further investigation of its role as an adjuvant to standard disease-modifying therapies in MS in Phase III trials.
Stem cells
Neural stem-cell transplantation. The use of self-renewing, multipotent stem cells that are capable of in vivo functional differentiation might provide a means for oligodendrocyte and myelin regeneration in MS. As axonal loss and functional disability correlate, this strategy could theoretically affect disease progression, not only via direct tissue repair but also by means of paracrine neuroprotective, angiogenic and anti-inflammatory effects, along with the expression of neurotrophic and survival factors that could stimulate endogenous progenitor cells.
It is known that endogenous oligodendrocyte progenitors are present in the adult brain237. In mice, transplanted human neural stem cells (NSCs) and embryonic-derived stem cells (ESCs) can differentiate into oligodendrocytes that have remyelinating potential238, 239. In acute EAE models, both ESC-derived and NSC-derived neurospheres, injected either intrathecally or systemically, are able to penetrate deep into the brain parenchyma240, 241. Animals that had received these transplants display clinical and histopathological disease attenuation, as reflected by higher numbers of endogenous remyelinating oligodendrocytes, decreased astrogliosis, and less prominent axonal loss and inflammation. Although stem-cell transplantation was initially attempted with a purely direct reparative goal, its beneficial effect is probably the result of various partially understood mechanisms. The transplanted NSCs seem to follow white-matter tracts, and preferentially target areas of inflammation (persisting as proliferating undifferentiated progenitors that secrete anti-inflammatory cytokines) and neurodegeneration (where they differentiate into mature oligodendrocyte cells with reparative/axonal remyelinating potential). Interestingly, in this model, stem cells may exert an additional anti-inflammatory effect by inducing apoptosis in invading CNS autoreactive T cells242. Furthermore, it has recently been suggested that NSCs may additionally reduce brain inflammation through downregulation of lymphocytic infiltration, a decrease in expression of intracellular adhesion molecule 1 (ICAM1) and lymphocyte function-associated antigen 1 (LFA1), and upregulation of suppressor regulatory T cells243. In addition to fundamental ethical and safety considerations, the use of ESCs faces a number of technical challenges, such as achieving optimal in vivo proliferation and appropriate differentiation into functional neurons and glia, while avoiding potential tumorigenicity. MS in particular is a multifocal disease; therefore stem cells administered systemically, rather than intrathecally, should be capable of entering the CNS and preferentially migrate to abnormal/lesional areas. In this regard, one could presume that blood–brain barrier leakage, due to ongoing CNS inflammation, would facilitate CNS entry and targeted migration of the stem cells. On the other hand, effective control of detrimental ongoing inflammation may be necessary in order to avoid stem-cell transplantation failure. Although exciting advances have been made, CNS reconstruction by stem-cell transplantation remains an experimental but promising therapeutic approach.
Haematopoietic stem-cell transplantation. A complete 'reset' of the immune system during the inflammatory stage of MS, before irreversible disease progression and disability ensue, might prove effective in disease control. With this goal in mind, autologous haematopoietic stem-cell transplantation (HSCT) has been investigated in demyelinating disease.
When performed in the relapsing phase, this therapy achieved cure or remission in acute EAE models244. However, this approach did not initially translate into a benefit in MS, as patients who had received autologous HSCTs displayed ongoing active demyelination and axonal loss, which correlated with brain atrophy and disease progression, despite significant suppression of lymphocytic inflammation245, 246, 247, 248, 249, 250. It should be noted that treated patients suffered from advanced disease, with likely neurodegenerative predominance. In this case, activated macrophages and microglia could be responsible for the observed ongoing CNS damage, independent of inflammation251.
Inherent to the concomitant high-dose immunosuppressive therapy in autologous HSCT is the significant risk for serious infections and transplant-related mortality. In addition, the progressive brain atrophy observed in patients who had received transplants may relate to myeloablative chemotoxicity252. The use of modern non-myeloablative HSCT protocols, with lymphoablative agents that have lower toxic effects seems logically advantageous from a risk/benefit standpoint. In view of all the facts, young patients with aggressive (predominantly inflammatory) or refractory disease but low disability therefore appear to be the best candidates to potentially benefit from this approach. Further investigation on the safety, effectiveness and long-term durability of this therapy is currently taking place in the Multiple Sclerosis International Stem Cell trial.
Conclusions
Traditional views of MS have radically changed within the past decade. We are now starting to understand that this heterogeneous disease is a constitutive diffuse syndrome in which compartmentalized CNS inflammation and neurodegeneration are present even at early stages and are maintained as a result of multilevel immune pathogenic changes. In particular, innate immune system dysfunction may have a key role in disease maintenance and progression.
The poor efficacy of approved disease-modifying therapies in disease progression may relate in part to at least two factors. First, these agents are known to affect adaptive autoinflammatory effectors, which predominate in the relapsing stages, but their contribution to restoring innate immune dysfunction appears to be limited. Second, they do not significantly affect neurodegeneration, which closely correlates with functional disability in patients with progressive MS.
The complexity of MS pathogenesis challenges us to keep open minds and to develop future plans of attack from multiple fronts (which might also include those related to potential environmental factors; see Box 3). A shift towards the development of drugs that specifically target the immune dysfunctions identified at several levels so far in different MS stages, and their use in combinations rationally aimed at maximizing synergistic benefits, should prove a positive advance in MS treatment. In this regard, several mAbs and some oral agents seem to be novel promising therapeutic options. In addition, the design of agents that are capable of preserving neural function and, ideally, preventing neurodegeneration, should also be the focus of our efforts to defeat MS, an adversary with many faces.
4
1 integrin in EAE.
