"Treatment Targets in Multiple Sclerosis, The ends and the means”

XIII European Charcot Foundation Lecture

STEINMAN, Stanford, USA: New targets of treatment in multiple sclerosis

High-technology approaches have helped us in understanding diseased tissue in MS. And since this is the year of some big breakthroughs in genomics for MS, it is interesting to compare our learnings from studying diseased tissue with genomic studies. Among the molecules that play an important role in MS but that we haven't heard about in genomic studies are osteopontin, alpha-B crystallin and two members of the coagulation cascade, the inhibitor of protein C and tissue factor, and the renin-angiotensin system.

It has been argued in the Annals of Neurology that EAE is a misleading model for MS and that it has failed to point towards a meaningful therapy or therapeutic approach in MS. Instead, the EAE model has proven itself a remarkably useful tool for aiding research in MS over the past 77 years. It is true that EAE does not represent MS in animals, but it can be used like a Petri dish is used by a microbiologist to elucidate the function of genes that have mutated. The EAE model has helped us to understand the biology of new and unexpected targets that may present themselves in MS. This has resulted directly in the development of three therapies, approved for use in MS: glatiramer acetate, mitoxantrone and natalizumab.

In many ways MS is an ‘outside in' disease. There is a big inflammatory component which in large part is mediated by lymphocytes that come from outside the CNS, through the vasculature and the extracellular matrix (by metalloproteasis). Once in the brain there is a complex interplay between microglial cells, macrophages and a variety of lymphocytes, including T- and B-cells. The exact function of complement and other aspects of innate immunity need to be placed into context. At this moment we are very interested in targets. We are looking at cellular targets like the B-cells and the blocking of various cytokines (gIFN, IL23 and osteopontin). Also we want to understand more about the working of drugs like glatiramer. Is it working at the level of antigen presenting cells, is it antigen specific or not?

The 5 ‘omics' (genomics, transcriptomics, glycomics, lipodomics and proteomics) are of vital importance for our knowledge of MS. In 2007, in addition to what we already knew about the role of HLA in the susceptibility to MS, the receptors for IL2R and IL7R were found to be associated with MS susceptibility. However, these two new genes have a much lower impact than HLA. As a group they are the foundation for understanding the genomic basis of the disease. But genomics are only one part of the story. What have we learned from studying the gene transcripts within the lesions, what are we learning from studying glycomics (i.e. carbohydrates found in the lesion), lipodomics (lipids found in the lesion) and proteomics (proteins found in the lesion)? This search has led us to some new insights in the pathophysiology of MS, which invariably leads to better treatments in predictive and personalized medicine.

In 2001 osteopontin was identified as number 5 on a long list of MS specific genes. A list that was headed by alpha-B crystallin. Since then efforts have been made to clearify the pathophysiologic role of osteopontin in MS. Osteopontin is part of a cluster of sibling genes, small integrin binding proteins that are clustered together on chromosome 4q21. This cluster of genes has one common operational function which has to do with the formation of teeth. Since it has become clear that humans have fewer genes than expected a decade ago, we are realizing that genes play modular functions. In one context a gene like osteopontin may be profoundly important in forming teeth, whereas in another context it may be important in controlling inflammation. This illustrates that by giving a name to a gene we can become easily distracted and it can constrain our thinking about other functions that they may have. Also giving amusing names (like ‘fuki sushi') to specific genes that later turn out to be of vital importance in human diseases, may cause embarrassment to physicians having to explain to their patients that their condition is a result of a mutation is a gene with an amusing name. In order to avoid both distraction and embarrassment it will be better to identify genes by giving them numbers and letters.

Osteopontin is a small integrin binding protein that binds a4b1 integrin and also has a thrombin cleavage site that binds fibronactin. Osteopontin is widely present in MS lesions as in a lot of other inflammatory brain diseases. In EAE animals with the osteopontin knocked out the disease is much milder, animals do not die and also the course of the disease is affected. From a publication in 2003 we know that elevated osteopontin plasma levels are detectable even prior to the actual relapse event in RRMS patients. Osteopontin is complicated to measure in the plasma because it is bound to factor H. Factor H is an inhibitor of the complement cascade. It's mutations are important for a number of diseases, including macular degeneration. And so there is a curious connection between de complement cascade and osteopontin via factor H. A recent animal study has determined a causal relation between elevated osteopontin levels and the onset of a relapse. One way to explain this causal relation is by erasing proinflammatory cytokines, including gIFN and IL17. Another way is that osteopontin affects the apoptosis, whereby osteopontin seems to act as a pro-survival molecule preventing this programmed cell death. By keeping activated T-cells alive, osteopontin allows these cells to keep delivering their pathologic blows.

This presents a possible new target in the treatment of RRMS. By closely following a patient's osteopontin plasma levels, a relapse could be prevented by administering a monoclonal antibody against osteopontin before the relapse actually occurs. At this moment these methods seem rather far fetched, but these are the things that will happen in 12 years time when osteopontin plasma levels can be measured in a nanotube that persons can carry in their wallets.

Alpha-B crystallin has been a focus in MS since 1995, when a brilliant experiment was performed by people in Leiden (The Netherlands). This experiment exemplifies what the MS tissue can teach us about the details of immune response. The results from this experiment unexpectedly pointed towards a fraction in the myelin sheath of MS patients that drive T-cells to produce gIFN the most: alpha-B crystallin. After a series of experiments that were publicized in 2007 it became clear that alpha-B crystallin has a protective role and is therapeutic in autoimmune demyelination. Alpha-B crystallin is a member of the small heat shock family of proteins (e.g. HSP27/HSPB1). Its name is derived from the fact that it constitutes approximately 35% of the proteins in the lens and is one of the major structural protein components that produce the necessary refractive index. High levels of alpha-B crystallin are found in the lenses of the eyes, adult heart and skeletal muscle; lower expression is found in kidney, lung, CNS (astrocytes, oligodendrocytes) and liver. Alpha-B crystalline had multiple functions. It is anti-apoptotic and protects cells from various stresses (thermal, osmotic, oxidative stress, TNF, myocardial ischemia and reperfusion, etc.). Its protection from apoptosis is via inhibition of caspase-3 activation.

In EAE animals with alpha-B crystallin knocked out the disease is worse, there is a higher proliferation of lymphocytes and there is much more production of gIFN and IL17. Also there is a higher amount of apoptosis of glial cells and upregulation of pro-inflammatory parameters in astrocytes, T-cells and macrophages.

Alpha-B crystallin is turned on by the MS brain as a protective molecule to diminish the inflammatory response. In fact it acts as a potent molecular check or brake on breakdown inflammation. In MS the immune system reacts out of context to alpha-B crystallin and assaults this ‘guardian of the brain' with a very strong immune response against it. When administering recombinant alpha-B crystallin i.v. to EAE mice, the disease gets better and the pro-inflammatory cytokines are diminished and the anti-inflammatory cytokines rise. Potentially recombinant alpha-B crystallin might one day be a therapeutic by itself or at least a way of turning off a counterproductive immune response against one of the brains guardians.

The human body has a number of preformed proteins that guard and protect it instantly against external assaults (e.g. a bite from an animal causing rabies). These proteins are readily available and do not depend on an evolutionary, selective, adaptive process like the formation of IgM and IgG. Among these proteins are alpha-B crystallin, alpha-1 antitrypsin, members of the small heat shock family, factor H and activated protein C.

Tissue factor and the inhibitor of activated protein C (PAI-3) are both members of the clotting cascade found in chronic active MS plaques. Both are pro-coagulatory molecules. There is always a balance between fibrinolysis and coagulation. It has been understood that inflammation somehow favors a pro-coagulatory state and by giving anti-coagulants one might attenuate autoimmune neuroinflammation. This appears to be a promising approach based upon a series of EAE studies using Refludan and recombinant activated protein C. So some day we might think about treating MS with drugs that modulate the clotting cascade. We can even engineer these molecules so that they affect pro-inflammatory signaling without affecting clotting itself. This could be new part of the therapeutic arsenal.

Angiotensin converting enzyme (ACE) is another molecule that was found from the proteomic studies in active MS lesions. It has been known from literature that there is an increased ACE activity in the CSF op MS patients. The EAE brain is filled with angiotensin receptors. In EAE experiments the blockade of ACE (e.g. by lisinopril) turns off the pro-inflammatory cytokines and gets milder EAE. And so lisinopril might be a really inexpensive drug in our arsenal of treating MS.

In the past several attempts have been made to create a prophylactic or therapeutic vaccine that would be antigen specific to turn off MS. From these attempts we have now learned that the innate immune system recognizes not only the lipids, proteins and sugars, but it also recognizes the DNA of microbes. There are DNA sensors in all the cells of our body that respond to double stranded DNA and can sense that a microbial pro-inflammatory DNA motive (called CPG) is around, driving a very potent inflammatory response.

BHT-3009 is a newly developed double stranded plasmid DNA-vaccine. Its phase IIb results were presented recently during the 2007 ECTRIMS conference. Primary endpoint was the median 4-week rate of new GD+ lesions after 28 – 48 weeks of treatment with BHT-3009 (two doses) or placebo. The lower dose arm (0,5 mg) showed a reduction of 50% on this primary endpoint versus placebo. Also there was a 67% reduction versus placebo in the 0,5 mg group in the total number of new MRI Gd+ lesions. From a predefined subgroup of 80 patients that underwent intensive immunologic studies, we learned that patients with the highest levels of anti-MBP antibody activity in the CSF had the highest reduction rate (61%) in number of Gd+ lesions. The higher the anti-MBP level in the CSF, the higher the reduction rate was in MRI-activity. No differences were found in relapse rates (0,4) between either groups. This result could have been due to the inclusion of quiet MS-patients (5 or less Gd+ lesions at screening) in this trial. Therefore the next trial will also be open to patients with higher numbers of Gd+ lesions.

In 2020 MS therapy will centre around the following 4 P's: Precise, Predictive, Personalized and at a realistic Price. There will be precise antigen specific therapy which will target adaptive responses against constitutive myelin targets and induced myelin targets. Large scale blockade of a whole category of cells like B-cells, or widespread impairment of a process so central to health and well being, like the ability of lymphocytes to home, will only be undertaken as a ‘second line' and will only be used ‘short term' to restore the opportunity for more precise and less risky therapy. New therapeutics to target the innate immune system will surely be employed including new generations of modulators of type 2 monocytes, next generation glatiramer, TLR antagonists, and agonists of protective heat shock proteins, like the crystallins.

New therapies will have a predictive arm associated with their use. We will know a priori who will likely respond to a therapy and who will not, but also who will likely be harmed by a therapy and who will not.

Therapy will be personalized based on the person's genomic background and exactly what their adaptive immune system is targeting. Nanoscale diagnostics will be invented to do this by the next generation of scientists. Large scale analysis will be done in a nanometer world.

Finally pricing is going to be a dominant factor of our consideration in developing new MS treatments. They will have to become available at a realistic and tolerable price.

Adjunctive therapy will target key inflammatory pathways: coagulation cascade, osteopontin, renin-angiotensin system, isoprenoids including Vitamin D and kynurenines. Therapeutics in 2020 will not resemble the "one size fits all” therapies of the late 20th and early 21st century.

I. WHAT ARE THE ENDS?

Session 1. Targets for treatment of multiple sclerosis

EDAN, Rennes, France: What is the target of multiple sclerosis treatment: relapse or progression?

Several well defined trials have demonstrated real impact on focal inflammatory lesions of the brain, thus reducing the number of relapses. Some well designed trials have demonstrated impact both on relapse rates and disability progression. For example, the BENEFIT trial demonstrated an impact of 40% reduction in relapse rate while reducing disability progression with 40% in a selected population with a clinically isolated syndrome. Similarly a reduction was found in the AFFIRM trial with Tysabri, showing a reduction in both relapse rate and disability progression. When using mitoxantrone as an induction treatment before IFNb, a clear impact was demonstrated both on reduction in relapse rate and in disability progression. But there are also trials that have demonstrated a reduction in relapse rate, but were unable to detect an impact on disability progression. This was especially the case in the population of SPMS patients. For clinicians this seemed a very paradoxal situation. For how can we explain when with the same medication and the same disease a clear clinical impact on disability progression is found in some trials, but not in other trials? The key question is what is the relationship between relapses and disability progression?

It has been found that the clinical course of MS both in PPMS and SPMS depends only on age. The younger the age at clinical onset of MS, the younger the age at assignment of irreversible disability milestones. Also the age at disability milestones (EDSS 6) is substantially influenced by the type of the initial course of MS, being older in RRMS than in PPMS. Both age at onset and relapses influence disability progression only at the early stage of RRMS (EDSS 0 - 3). In the late stage of MS (EDSS 3 - 6) age does not influence the disability progression. And so age at onset of RRMS is a prognostic factor for early disability progression. For patients younger than 20 years of age it takes an average of 14 years to reach EDSS 3. In patients older than 50 it takes an average of 3 years to reach EDSS 3. And therefore it is assumed that ageing related processes may play a role in the early stage of RRMS. Remarkably there is no time difference in the later stage of the disease between age groups to go from EDSS 3 to EDSS 6.

Earlier studies had found that the numbers of relapses in the first 2 years after onset of the disease influence the rapidity of progression from clinical onset to EDSS 6. It has been shown that it only influences the rapidity of going from onset to EDSS 3, but not on the period from EDSS 3 to EDSS 6.

So the general situation concerning the relationship between relapse and disability progression can be summarized as follows. MS is a 2-staged disorder. The first stage is called the focal inflammation stage (EDSS 0 – EDSS 3). This stage is influenced by focal inflammation and age. The second stage, called the neurodegenerative diffuse inflammation stage (EDSS 3 – EDSS 6), is independent of focal inflammation and age. This stage is quite homogenous in the development time needed to reach EDSS 6.

In conclusion it can be stated that from the therapeutic experience of the last 5 years and also from revisited epidemiological data, the hypothesis has emerged that available MS treatments might significantly target both relapse and disability progression but only within a narrow therapeutic window in the early stage in the RRMS population. So the major message is that it is important to start treatment before patients reach the disability stage of EDSS 3-4.

BRück, Göttingen, Germany: What does pathology suggest on relapses and progression

The main pathological event occurring in a relapse is active ongoing demyelination. This is associated with inflammation and acute axonal damage occurring in an acute lesion within a relapse. The pathophysiological mechanism leading to the clinically acute relapse is probably conduction block due to inflammatory mediators that are present within the lesion.

Active demyelination is a major feature in the relapse period, also in a secondary progressive period of acute relapse. There are a number of candidates that have a possible pathological correlation with disease progression. Slowly expanding pre-existing lesions may be one of the major correlates as opposed to newly developing lesions which are typical for the relapsing remitting course. Other possible correlations are persistent microglial activation, compartmentalized inflammation, remyelination failure , axonal/neuronal loss, cortical/gray matter demyelination, B cell/antibody involvement, endothelial abnormalities and changes in normal appearing white matter

Slowly expanding pre-existing lesions (progressive plaques) in which ongoing myelin breakdown occurs in the absence of acute inflammation, contribute to disease progression in cases of SPMS. Microglial cells digesting myelin in the pre-existing plaques may be the major driver in progressive MS-lesions. Therefore microglia may be an interesting therapeutic target in the progressive disease stage.

Remyelination failure does occur in chronic MS. About 40% of the chronic MS plaques show signs of remyelination. Remyelination is very prominent in a subpopulation of MS patients. It seems also that remyelination is restricted to a subpopulation of MS patients. The question is still open whether the population of MS patients that does not show remyelination has a worse prognosis and a more extensive secondary progression of MS. It may well be that patients showing remyelination have a better prognosis.

Axonal loss is certainly a major contributor to progressive MS. In the chronic disease stage about 60% of the axons are lost. In animal studies neurological impairment has been found to correlate with axonal loss in chronic EAE. If axonal loss transcends a certain threshold (often about 30%) a permanent clinical deficit is established. The same may occur in human MS patients. When a certain threshold of axonal loss is reached a patient may enter a secondary progressive disease course. Remyelination protects axons. Patients with remyelated axons are better protected against axonal loss and have a better prognosis and do not enter the secondary progressive disease stage as early as non-remyelinating patients.

Cortical demyelination is much more extensive in PPMS and SPMS than in RRMS. The involvement of the cortex may be a phenomenon of secondary progression. There seems to be an accumulation of meningeal B-cells, especially in a subgroup of patients with SPMS. This points towards a role of antibodies in secondary progression.

Conclusion:

The exact pathological correlate of disease progression in MS is not yet defined. Although we have made a number of steps forward in the last few years. There are a number of candidates that could be especially interesting for developing therapeutic targets. Different factors may contribute to progression including:

· progression of existing lesions of compartmentalized inflammation

· remyelination failure (induction of remyelination: with the knowledge that in the chronic lesions OPC's are present that could be a target)

· axonal/neuronal loss (microglial cells are an interesting target since they are probably the main mediators of ongoing damage in the progressive disease stage).

· cortical involvement

· antibodies

FAZEKAS, Graz, Austria: MRI: What should we measure?

MRI measurement has certainly played a major role in drug development so far. Still there remains a need for improvement. In the context of drug development we should attempt with MRI to measure specific pathophysiological processes which correlate closely with relevant clinical outcomes and respond to distinct therapeutic interventions in a sensitive and reproducible manner.

Table: Usefulness of currently available MRI measurement techniques

Active contrast enhancing lesions with Gadolinium (Gd) are already quite familiar in use. This indicates the break down of the blood brain barrier. The occurrence of new (focal) T2 lesions is consistent with new areas of MS related tissue damage. T2 lesion load reflects some accumulation of gross tissue damage. T1 lesion load is indicative of the proportion of more severe parenchymal destruction. Brain/spinal cord volume changes are 

indicative of alterations in tissue integrity (including cell loss / increase) and composition (including water). They are frequently associated with neurodegeneration. Other quantitative MRI techniques (such as MIT, DWI/DTI and MRS) look quite promising, but are not yet ready to be applied.

The current situation with regard to the application of MRI techniques in MS can be described as follows. We have an established set of (‘conventional') MRI techniques that provide outcome variables that are reasonably well understood as to their reflection of distinct pathophysiological processes of MS. Guidelines and recommendations have been developed to assure their reliability, reproducibility and use in a multicenter setting. The correlation of these MRI outcomes with clinical variables is limited, although they may be better than assumed. They have served to support efficacy in all pivotal trials of currently approved drugs. Also the established outcome variables are heavily weighted towards signs and consequences of inflammation and may not be appropriate for other treatment targets (e.g. neuroprotection). MRI outcome variables are especially important for phase II studies. This is increasingly recognized and accepted by regulatory authorities. Evaluation of and search for other MRI outcome measures has to continue (including provision for their applicability in multicenter trials). To understand their response to treatment exploratory MRI variables (e.g. MIT, DTI, MRS) need to be included in all drug trials in order to gain a better understanding about how they can be interpreted in view of the pathophysiological processes of MS.

II. WHAT ARE THE MEANS?

Session 2. Target 1: B and T-cell activities in multiple sclerosis. The peripheral part of the immune system

HOHLFELD, Munich, Germany: B and T-cell activity in multiple sclerosis

When looking at T-cells in MS a number of considerations can be made. Traditionally investigators of T-cells in MS have taken one of two different, but complimentary, approaches both pioneered around the late 80's. One is called the candidate antigen approach in which immune cell responses against candidate antigens are investigated. Antigens identified in animal models (e.g. MBP, MOG, PLP) are taken into the human system. Investigators have extensively characterized human T-cell response against all candidate antigens that have been derived from animal experiments in EAE. The second is called the T-cell repertoire approach (TCR). In this approach antigen receptor expression is investigated in T-cells which infiltrate MS lesions and use identified receptors for the screening of antigen libraries.

From the candidate antigen approach a list of potential target antigens of T-cells have been derived that play a role in pathogenesis of MS. These include MBP, MOG, PLP, alpha-B-crystallin, non-myelin auto-antigens (S-100) and foreign (viral?) antigens in any possible combination. The problem is that T-cell responses against all mentioned auto-antigens can be demonstrated also in healthy people. It is virtually impossible to demonstrate any differences between the auto-immune, auto-reactive T-cells isolated in healthy humans and those in MS patients. This is a fundamental hurdle that has not yet been overcome.

From the TCR approach it has become clear that CD8⁺ T-cells play a distinct role in the pathogenesis of MS. They are present in different compartments (CNS, CSF, blood) and persist for a long time (even more than 5 years). CD8⁺ T-cells show evidence for recent activation and antigen driven selection. This means that they are recruited into the brain by some particular unknown antigen. And finally expanded T-cell clones ‘pervade' distinct anatomical regions including widely separated lesions in the normal appearing white matter (NAWM).

When looking at B-cells and antibodies in MS an extensive list of potential targets can be given, including MOG, aquaporin-4, neurofascin, alpha-B-crystallin and many others. Many studies have shown that the few B-cells that are present in the CNS or CSF of MS patients are clonally restricted or expanded (at the transcript level contained in cells). Likewise, the antibodies present in CSF are oligoclonal (at the protein level). The target antigen(s) of these oligoclonal bands (OCB's) remain unknown. Whether the OCB's are the clue to the mystery of MS is still debatable. Recent study has confirmed that B-cells found in the CSF are one of the sources of the OCB's.

We can be extremely optimistic in view of the technological advances made in the last decade. These will allow us to make new discoveries leading to new concepts of pathogenesis and disease heterogeneity, and also to new treatments.

OCHI, Fukuoka, Japan: New immunosuppressive approaches

Immune tolerance is a state of immune system unresponsiveness to an antigen. Autoimmune diseases arise from imbalance between the auto-regulatory immune system and pathogenic auto-reactivity. Strategies to induce immune tolerance are being developed for the treatment of autoimmunity. The mucosal tolerance approach involves mucosal, oral and nasal administration of auto-antigens designed to induce regulatory cells. A second approach is parenteral administration of an anti-CD3 antibody, which has been shown to be efficacious in animal models of autoimmunity and in humans with type 1 diabetes. The first demonstration that anti-CD3 could affect autoimmune diabetes was reported in the mid 1990's using NOD mice, which is a mouse model of type 1 diabetes. 80% of NOD mice develop diabetes spontaneously. When given IV or IP, anti-CD3 induces disease remission within 2 to 4 weeks of treatment. This effect lasts for 8 to 10 months and is thought to be islet antigen-specific. T-cell depletion is one of several possible mechanisms. IV injection of anti-CD3 induces depletion of T-cells (40-50%), but it is transient and cannot explain the long lasting clinical effects. TCR down-modulation is also transient and only occurs during the treatment phase. Recently, it was reported that TGF-beta dependent CD4+CD25+ regulatory cells were induced by IV anti-CD3 and mediated disease remission in NOD mice. Data have shown that tolerance induced by anti-CD3 depends on TGF-beta producing regulatory T-cells in NOD mice.

Immune modulation of autoimmunity by mucosal administration of auto-antigens designed to induce regulatory T-cells was investigated. It was found that (1) TGF-beta secreting regulatory cells are generated during oral tolerance, (2) oral OVA induces regulatory T-cells in OVA-transgenic mice, (3) even small peptides and cytokines are biologically active when given orally. (4) In addition, it was recently reported that signalling through IV anti-CD3 affects T-cell function and induces TGF-b-dependent CD25+ regulatory cells in NOD mice. Given this background it was hypothesized that oral anti-CD3 would induce immunologic tolerance and would not have potential side effects such as cytokine release syndrome.

So far, recent studies have lead to the following main conclusions:

· Anti-CD3 is biologically active when given orally

· Oral anti-CD3 antibody suppresses EAE both in preventive and therapeutic paradigms

· CD4+CD25-LAP+ Treg cells play a role in the immunomodulatory effects of oral anti-CD3 antibodies

· CD4+CD25-LAP+ Treg cells suppress via a TGF-b dependent mechanism.

· Mucosal anti-CD3 is effective in animal models of diabetes and lupus

ANTEL, Montreal, Canada: The first data on the rituximab trial - Proof of concept of B-cell involvement in RRMS

The followed strategy behind the design of the trials performed with rituximab challenge us in defining the clinical aspects of MS that need to be evaluated and the best way to monitor in order to find out more about the principles of how a specific agent works. The available data are based on phase I and II trials conducted with rituximab in RRMS. Rituximab is a chimeric mouse human monoclonal antibody that targets CD20 expressed on cells of the B-cell lineage from pre-B-cells up to plasma cells. CD20 is not expressed on mature plasma cells.

The main entry criteria for the phase II trial were RRMS, 18-55 years old, 1 relapse in the past year and EDSS ranging 0-5. Patients received two dosages of rituximab or placebo early in the trial and were then followed up to 48 weeks. 104 patients were actually randomized at a ratio of 2:1, either on rituximab (69 patients) or placebo (35 patients). After 48 weeks 21 (60%) patients on placebo and 58 (84%) on rituximab completed the trial. Most drop-outs left the trial after a 24 weeks period. After 48 weeks 14 (40%) patients on placebo and 14 (20%) patients on rituximab had experienced a relapse. From this an annualized relapse rate was calculated for the placebo group (0,719) and the rituximab group (0,374) after 48 weeks. Total Gd lesion count in MRI analysis was a primary endpoint. 18 (51%) patients on placebo and 53 (80%) patients had no Gd lesions. The mean number of Gd lesions in the placebo group was 5,5 versus 0,5 in the rituximab group. With regards to the CD19 counts (reflecting the B-cell measurement) there was a clear drop as a result of rituximab.

It will be a recurrent theme from a clinical perspective that with the use of these type of agents you will have a considerable number of patients with suppressed immune activity for a prolonged period of time. It remains to be seen how the clinical community will deal with this in view of issues of reversibility. Also it needs to be established whether the properties of the immune system have changed once it has been reconstituted and whether this type of approach provides a prolonged (and hopefully) beneficial effect.

At the end of the phase II trial we seem to have an effect on RRMS based upon both clinical and MRI criteria. Issues that need to be further reviewed with regard to rituximab are:

· What is the basis of rapid clinical/MRI response without change in serum or CSF (based on previous trials) immunoglobulin levels? From open label studies with neuromyelitis optica and rituximab there have been claims of reduction of the course of the disease without a change in antibody titers.

· Is recovery of B-cell lineage cells associated with recurrence of clinical disease activity (i.e. does not induce immunologic tolerance)?

· What is the risk of repeated therapy? How many times are we prepared to treat with this kind of approach?

· What is the effect in progressive disease? A major trial in PPMS is ongoing.

LUTTEROTTI, Hamburg, Germany: Cell based antigen specific approach to induce tolerance in multiple sclerosis

The down regulation of immune response in autoimmune diseases through the induction of antigen specific tolerance is the ultimate aim of immune therapy. Several strategies have been successfully used in recent years and are currently being clinically developed, such as DNA vaccination, intravenous administration of soluble MBP and TCR vaccination. These all have different approaches in inducing antigen-specific immune tolerance, either by inducing anergy, activation induced cell death/apoptosis, bystander suppression (Th2 shift) or induction of regulatory cells (Tr1, Treg).

The approach that is currently under investigation in Hamburg is to induce tolerance by peptide coupled cells. Peripheral blood mononuclear cells (PBMC's) obtained by aphoresis from a MS patient are incubated ex vivo with peptides which are aimed to induce tolerance. The PBMC's covered with peptides are then infused in an autologous way into the patient. When thinking of an antigen-specific therapy it is good to realise that there may be four immunodominant peptides that play a role in MS. There is evidence of epitope spreading in RRMS, meaning that with each relapse the immunodominant peptide and the immunospecific response is diversifying. This means that any antigen-specific therapy in MS should involve the induction of tolerance against several immunodominant peptides. This approach has been shown in animal studies with established EAE mice to have a therapeutic effect by abrogating epitope spreading when administering immunodominant peptides at the peak of the first relapse.

The assumptions based on animal data collected in recent years are that the peptide coupled cell tolerance approach is effective for the prevention and treatment of relapsing EAE and that it is exquisitely antigen specific. Also it can simultaneously induce tolerance to multiple peptides and it is more effective than intravenous or oral administration of soluble peptides. Peptide coupled cell tolerance has been shown to down regulate antigen specific response in human T-cell clone in vitro. The mode of action presumably involves both direct and indirect mechanisms (by apoptosis). According to current knowledge tolerization by peptide pulsed cells is safe.

Next step is to start an open-label, single center, baseline-to-treatment cross-over phase IIa clinical trial. This trial is designed to:

· assess the safety and tolerability of antigen-specific tolerization with peptide-pulsed PBL in RRMS

· to examine the efficacy of tolerization with peptide-pulsed PBL in reducing the inflammatory activity in the central nervous system.

· to investigate the mode of action of peptide-pulsed ECDI-fixed PBL on inflammatory activity in MS.

The peptides used in this trial include four MBP peptides (13-32,, 83-99, 111-129, 146-170), one PLP (139-154) and 2 MOG peptides (1-20, 35-55).

SELMAJ, Lodz, Poland: The cytokine network in multiple sclerosis

Cytokines are biologically active polypeptides released at sites of inflammation by a variety of cell types ranging from fibroblasts, astrocytes, macrophages and T-cells. Cytokines have a wide variety of effects on the immune system ranging from activation to suppression. There is much literature on the presence of cytokines in MS lesions. Both Th1 (e.g. IFNg, IL-12 and IL-6) and Th2 (e.g. IL-4 and IL-10) cytokines are heavily upregulated in various cell types in MS lesions. EAE (and probably also MS) is traditionally a Th1 mediated disease.

Th1 cytokines can contribute to encephaliogenecity and EAE on several levels. In the early phase they can contribute to immune cells homing to the CNS. TNF can induce injury and death of oligodendrocytes. Also cytokines can contribute to inflammatory neurodegeneration in combination with CD8⁺ lymphocytes, microphages, microglia, nitric oxide, MMP and fibrine and they can be involved in axonal damage in MS.

In recent years the Th1/Th2 paradigm in EAE was elaborated by Th17-cells as a third category. Th17-cells are supposed to create the strongest immunogenic endorsement of EAE. Th17-cells are generated from (naive) Th0 cells by dual stimulation of a regulatory cytokine TGFb and IL-6. IL-17 secreting Th17-cells are proactivated by IL-23 and downregulated by IFNg, IL-4, IL-2 and IL-27. However, this hypothesis of Th17-cell induction was challenged by a recent article in Nature Immunology, which states that TGFb and IL-6 not only drive the T-cell production of IL-17 but also IL-10, thus restraining Th17-cell mediated pathology. This means there still remains work to be done in view of the Th1/Th2 paradigm to determine the role of Th17-cells in MS pathology.

The development and function of Th1-cells is stimulated by Notch molecules. It has been shown that both Notch1 (through IFNg) and Notch3 induce Th1-cells. Some experiments have been performed with gamma-secretase inhibitor, which inhibits both Notch1 and Notch3 expression. Cytokines are not only active in pro-immunogenic functions in MS but are also involved in regulatory functions. Cytokines control the function of immunoregulatory cells such as CD4+CD25+ Foxp3+T-cells, Tr1-cells and Th3-cells. Experience with cytokines as therapeutic targets has been collected in the treatment of rheumatoid arthritis. So far, treatments against TNF, IL-1, IL-6, IL-12/IL-23 and IL-17 are in various phases of clinical development or already available. Current available cytokine therapies in MS are IFNb and GA. Several new compounds, such as daclizumab directed against IL-2, show promising results in clinical development phases.

Cytokine therapy is a realistic approach and to some extend is already available. New therapies are underway and will become available in the near future.

COMI, Milan, Italy: New oral treatments in multiple sclerosis

In the near future we will have several better possibilities in hand in the treatment of MS, which we will have to choose from. For the clinician efficacy and safety are of the most importance. But we have to realise that from a patient's point of view tolerability is of the utmost importance because especially in the beginning of the disease the patient lacks the perception of efficacy and safety. So in the real world tolerability is the main factor, particularly in early treatment. Tolerability and quality of life are greatly affected by the administration route. Currently available MS treatments are all administered parenterally, while 22% of the general public suffers from needle fobia and 50-80% of patients treated with IFNb or GA have injection site reactions. Also tolerability affects adherence to treatment which is a key determinant of effectiveness. The following five oral MS therapies are currently in phase III.

Cladribine

Cladibrine is a purine nucleoside resistant to adenosine de-aminase. It is a preferential lymphocyte depleting agent, a specific cytoxic anti-lymphocyte and monocyte effect, currently used in the treatment of Hairy Cell Leukemia. Cladibrine is approved in 37 countries, including the US.

Cladibrine is selectively concentrated and active in lymphocytes. By this way it has a substantive biological effect on immunosuppression. An earlier trial investigating the effect of parenteral cladibrine (high and low dose) in progressive MS patients had shown no effect in disease progression as compared to placebo. However, a dramatic effect was seen with MRI on the reduction of inflammation (almost cancellation of Gd-enhancing lesions). Beside the lack of effect of parenteral cladibrine there had also been some haematological adverse effects (thrombocytopenia, purpura and anaemia) which caused the cancellation of the drug development in progressive MS. After 10 years the potential of cladibrine was reconsidered and an oral form was developed. Currently there are two major trials ongoing in an early phase of RRMS. The results from these trials are expected by the beginning of 2009.

Fingolimod (FTY720)

Native B- and T-cells regularly circulate between blood and lymphoid tissue in search of foreign antigens, a process requiring S1P/S1P receptor interaction. Fingolimod interacts with sphingosine 1-phosphate (S1P) receptor and traps circulating lymphocytes within the peripheral lymph nodes, thus preventing them to invade the CNS.

There are two mechanisms by which lymphocyte egress from lymph nodes is inhibited. Firstly a down-modulation of S1P1 receptors on T-cells and a reduced response of cells to the egress signal provided by S1P. Secondly a persistent signalling at S1P1 receptors on the lymph node endothelium to reinforce its barrier function, further reducing transmigration of lymphocytes from the lymph nodes.

The S1P receptor is part of a family with 5 different subtypes (S1P 1-5). Four of these subtypes are targeted by FTY720. The S1P1 receptor is relevant for the egress of lymphocytes from secondary lymphoid organs. Interestingly the other subtypes are mainly present on CNS cells (oligodendrocytes, neurons), which provides a potential of targeting FTY720 in order to activate ectocytes that may have a neuroprotective effect, modulate the blood brain barrier and activate oligodendrocytes to remyelinate after an attack. This provides a combination of effects that looks quite promising for future approaches in treating MS. The expectations of these effects of FTY720 are substantiated by preliminary evidence from in vitro and in vivo studies. A phase II trial showed a reduction of Gd-lesions with 50-80% and a 50% reduction in annualized relapse rate after 6 months. An extension trial has shown a continuation of these effects after a 2 years period. No major safety problems were reported. There are several studies ongoing on RRMS. Interestingly a trial with PPMS is planned in March 2008. This is an indication for which there is still no therapy and no experimental trials are ongoing.

Laquinimod

Laquinimod is an oral immunomodulatory agent structurally related to linomide, but with a better safety profile.

It reduces leukocyte infiltration in the CNS and induces a deviation of MBP-specific cells from a Th1 to Th2/Th3 pattern. Two phase II trials have already demonstrated a reduction of inflammation by 40-50% as measured by the number of Gd-lesions in MRI. The drug has shown to have an effect after 16 weeks which progresses over time as compared to placebo. Also a reduction of 40% in annualized relapse rare was reported on the higher dose of the drug (0,6 mg). After observing laquinimod in more than 350 patients no major safety problems were found, which is of particular importance in view of the history of linomide. Two phase III trials are ongoing in RRMS, one with a placebo control and one also with an active control using Avonex as a comparator.

BG-12

BG-12 (dimethyl fumarate) is a second generation of fumaric acid. The drug has both anti-inflammatory (by influencing expression of cytokines and adhesion molecules) and cytoprotective effects. BG-12 may ‘protect' oligodendrocytes and neurons from inflammatory and metabolic damage via activation of Nrf2 and of the anti-oxidant and metabolic defense mechanisms. In addition BG-12 may dampen inflammation and promote its resolution by interfering with NfkB-dependent inflammatory signalling.

A phase II trial has revealed a clear reduction of clinical and MRI (new Gd⁺ lesions) activity in RRMS (up to 70% after 24 weeks as compared to placebo). The reduced number of black holes suggests that the drug gives some kind of neuroprotection. BG-12 caused a drop of the annualized relapse rate by 32% as compared to placebo. BG-12 was generally safe and well tolerated. Again two phase III trials are ongoing in RRMS, one with two doses of BG-12 and a placebo control and one also with an active control as a comparator.

Teriflunomide

Teriflunomide is an active metabolite of leflunomide, which is an approved treatment for rheumatoid arthritis. It blocks the mitochondrial enzyme dihydro-orotate dehydrogenase and inhibits T and B cell proliferation. Teriflunomide interferes with the interaction of T-cells with antigen presenting cells.

Teriflunomide showed statistically significant improvement compared to placebo in the number of enhancing lesions, T2 lesions and burden of disease over 36 weeks. There is a trend in lowering annual relapse rate and fewer patients suffered a relapse. Also EDSS progression with teriflunomide (14 mg) was lower. The drug has an acceptable safety and tolerability. Serious AE's included elevated liver enzymes, hepatic dysfunction, heartburn, groin infection, gastritis and rash. There are three phase III trials ongoing.

All reviewed drugs are potentially a first line therapy in MS. The results of the phase III trials will have to show which will come out as winners.

Session 3. Target 2: The blood brain barrier and inflammation. The entry of inflammation into the brain.

ENGELHARDT, Bern, Switzerland: Adhesion molecules and chemokines. How to affect the blood brain barrier.

The blood brain barrier (also called the neurovascular unit) consists of a layer of endothelial cells of the blood vessels in the brain, its basement membrane, the perivascular space and a secondary basement membrane, the glia limitans, which is formed by astrocytes. Vast numbers of pericytes are embedded in the basement membrane. Astrocytes (by its foot processes) cover around 80% of the blood vessels in the CNS. Some neurons send their dendrites into the perivascular space and may be important for the maintenance of the barrier. The blood brain barrier (BBB) prohibits free diffusion of any molecules across the barrier. This means that immune cells need unique mechanisms to cross this barrier.

In general immune cells enter a tissue following a multistep cascade including adhesion and transmigration. Leucocytes have to slow down in the blood stream in order to interact with the vascular wall. They do this with adhesion molecules of the selectin family. These help them to slow down and roll along the endothelium cells. At a velocity of 40 microliters/sec it is able to detect chemokines which bind to a chemokine receptor on the leucocyte surface. These chemokine receptors belong to the G-protein coupled receptors. They send a G-protein signal through the cell to a second set of receptors called the integrins. Integrins have an alpha- and beta-chain on their surface in an inactive state. The chemokine signal wakes up the integrin causing it to bind its ligand to the endothelium. The ligand belongs to the Ig-superfamily. Only after clinging on to the endothelium these immune cells can start transmigrating across the endothelium. This is a step that is not quite well understood so far. With this knowledge one can ask which molecules are involved in the multistep cascade that help auto-aggressive cells to enter into the CNS parenchym. A combination of studies has shown that the T-cell interaction with the BBB is unique. In the first step of the multistep cascade a4-integrins seem to be able to mediate the initial rolling or capturing event of the T-cells. There is still debate about the chemokines that are involved in the interaction with α4-integrin to activate their conformation. As mentioned before, selectins and their ligands are normally involved in mediating the rolling and tethering of leucocytes in inflammated tissue. There is an upregulation of E- and P-selectin on the endothelial cells and of L-selectin on the leucocytes. α4-integrins are in some yet unknown way able to replace and mediate the interaction of selectins between endothelial cells and leucocytes.

Regarding the role of selectins and their ligands in EAE, it had already been shown by Kubes and Constantin that in inflamed brain vessels inflammatory cells roll via E- and P-selectin and their common ligand PSGL-1. However, recent studies with SJL and C57Bl/6 mice have shown that neither E-, P- and L-selectin nor PSGL-1 are required for the inflammatory cell trafficking across the BBB in EAE. The explanation may well be that the therapeutic success of targeting α4-integrins does not rely on the inhibition of inflammatory cell trafficking across the BBB.

Advanced study of the structure of menigeal and parenchymal vessels has shown that they differ in the way that leucocytes are recruited across the BBB. E- and P-selectin by PSGL-1 mediate the rolling of leucocytes only in the meningeal vessels. Parenchymal vessels are different from meningeal vessels (e.g. they lack astrocytic foot processes). Therefore it is suggested that the molecular mechanisms of T-cells and leucocytes recruitment across the meningeal vessels is not identical to the mechanisms in the vessels of the CNS parenchyma. In the parenchyma α4-integrins are predominantly important for T-cell recruitment and selectins do not play a role.

DAVENPORT, Cambridge, UK: Blocking α4 integrin, a small molecule approach

Integrins play a role in cell-cell and cell-matrix interactions. Of special interest for MS are α4b1 and α4b7. α4b1 has two active ligands, VCAM-1 and fibronectin. VCAM-1 is the ligand with the greatest affinity. Regarding α4b7 the ligand with the greatest affinity is MAdCAM, followed by VCAM-1 and fibronectin. In case of MS the combination of α4b1 and VCAM-1 is the most promising. It has not been possible yet to manufacture therapeutic antibodies that selectively attack the α4-chain, thus taking out α4b1 and α4b7. This also applies for small molecules. All available therapeutic options take out α4b1 and α4b7 in various degrees. α4-integrin antagonists inhibit the rolling and migration of inflammatory cells into the brain tissue. In MS this leads to prevention of inflammatory cells reaching the lesions in the brain.

Natalizumab was the first anti-α4 antibody. It inhibits both α4b1 (VLA-4) and α4b7. It was launched mid-2004 as a treatment for MS based upon its impressive monotherapy Phase III results (relative risk reduction for progression of disability of 42% over 2 years; relative risk reduction for occurrence of relapses of 59 % over 2 years). Beginning of 2005 natalizumab was linked with 3 fatal cases of progressive multifocal leukoencephalopathy (PML) and was voluntarily withdrawn from the market. PML is a disease associated with severe compromised immunity. At the time the FDA placed a clinical hold on all trials involving α4 inhibition while safety review was performed. In June 2006 natalizumab was subsequently relaunched with some restrictions and the clinical hold was lifted on trials involving α4 inhibition. α4 inhibition still is a promising clinical target.

Small molecules provide a number of benefits. First of all they are easier to use for patients as compared to antibodies that require infusion in either a hospital or surgery setting. Secondly, they washed out in shorter time (hours instead of months). Thirdly there is an opportunity with small molecules of different dosing for different patients, so called personalised medicine which is an area that is valued among clinicians. And also there is reduced cost of treatment. Biologicals tend to cost more to manufacture than small molecules and biologicals have more administration costs, especially related with clinics and doctors time. Including administration natalizumab treatment estimated costs may exceed $50.000 per year per patient.

Over the last 10 years more than 300 patents have been published by the pharmaceutical industry claiming α4-antagonists. This large number reflects the challenge that the pharmaceutical faces in discovering α4-antagonists. Apart from natalizumab, an α4 antibody already on the market, there are a number of α4b1/α4b7 antagonists currently in development for MS. These include GSK/Tanabe's Firategrast (GSK-683699) and UCB/Biogen's CDP323, both in phase II, and Encysive/Schering Plough's TBC-4746 in phase I.

Oral drug profile involves a delicate interplay of a number of factors, including potency to the target, absorption, metabolism and elimination. the problem with acid absorption is that carboxylic acid tends to ionise. The ionised form is not able to pass the GI tract, which means that only a small proportion (up to 1/10.000 depending on pH) of the drug will be absorbed. A pro-drug strategy (as is used in CDP323) can be used to overcome this problem and increase the drugs permeability in the GI tract. α4-antagonists are not cleared by phase I or phase II metabolism. Clearance of α4-antagonists is generally non-metabolically driven. The parent drug accumulates in the bile duct by a transporter driven process. Biliary elimination does complicate drug discovery for this is a relatively new area of research.

a4 research has spanned a 20 years period, which illustrates the challenges faced in discovering a4-antagonists. If efficacy of small molecules is comparable to that of the a4 antibody (natalizumab), they will represent an exciting new option for clinicians. CDP323's phase II results can be expected by the end of 2008.

ZIPP, Berlin, Germany: Blockade of chemokine signalling

We think that myelin-specific T-cells are the initiators of the inflammatory process in the brain and that these cells after activation and expansion are capable of transmigrating the BBB. Once they are activated locally in the brain, they can induce a damage cascade leading to demyelination and neuronal pathology. There are several therapies targeting to specific steps in the pathogenesis of MS. Our knowledge of chemokines and adhesion molecules is developing rapidly. It has become clear that chemokines perform a number of different tasks and are involved in inflammation. Different families of chemokines exist, such as the C-chemokines, the CC-chemokines, the CXC-chemokines and the CX₃C-chemokines. Of these the CC-chemokine family is the largest, with the CC-chemokine receptor-1 (CCR1) as one of its members.

Evidence for CCR1 involvement in MS pathogenesis consists of the following. Ligands of several chemokines (such as MIP-1α and RANTES) which are known to be proinflammatory in MS bind to the CCR1. CCR1 expression has been associated with newly infiltrating monocytes in MS lesions. The degree of leukocyte infiltration into the CNS during EAE was reduced upon treatment with a specific CCR1-antagonist. CCR1-deficient mice showed an attenuated EAE course. Also a small-molecule antagonist (BX 471) showed reduction in the severity of EAE and was able to delay heart transplant rejection. Moreover, a pilot study with CCR1-antagonist CP-481715 in a small group of patients with rheumatoid arthritis showed a trend towards clinical improvement and a reduction in the number of macrophages in the synovium.

However, a phase II clinical treatment trial with CCR1 antagonist ZK 811752 (Schering) among patients with RRMS performed between 2003 and 2005, has showed no difference in efficacy as compared to placebo. This tells us that the interference with the CCR1-pathway is not effective in MS. It also showed not to be effective in follow-up trials with rheumatoid arthritis or atopic dermatitis. Therefore the CCR1-principal that might have been as effective as blocking adhesion molecules has been proven not to be effective in the human system.

Despite this disappointing history of CCR1-antagonists, there are still numerous chemokine-receptor antagonists in development. With regard to MS CCR2-antagonism is being investigated in phase II (Merck and Millennium) and phase I (Incyte). CXCR4-antagonism is developed (phase III, AnorMED) in relation to stem-cell mobilization. These developments look quite interesting but we have to await the results of the clinical trials. It also shows that the chemokine system is very complicated and that much more clarification is needed.

STüVE, Dallas, USA: Risk benefit ratio of natalizumab

T-cell adhesion has been a major target of drug development over the last 15 years. The first molecule that was brought into clinical trials was approved for use in patients with RRMS. It was a monoclonal recombinant humanized antibody called natalizumab. Natalizumab binds to the α4 chain of VLA-4 which is expressed on all leucocytes (with the exception of neutrophils). This antibody was not designed to destroy its target cell, but rather to physically interfere with the interaction of VLA-4 with its natural ligands FN and VCAM-1 on the vascular endothelial cells and thus make it impossible for these cells to migrate to the peripheral brain tissues. There has always been concerns that keeping T-cells out of peripheral tissues might leave infections and malignancies unnoticed.

There is little doubt about the efficacy of natalizumab in RRMS. Two phase III trials showed significant effects with natalizumab both on primary and secondary outcome measures. Based on those two trials, natalizumab was approved in the US in November 2004. Three months later natalizumab was linked to 2 cases of progressive multifocal leukoencephalopathy (PML) and was voluntarily withdrawn from the market. PML is a disease that is typically seen only in individuals that undergo prolonged severe immunosuppression. PML is almost invariably fatal and unfortunately one of these patients died as a consequence of natalizumab therapy. One month later another PML-case was made public, a patient who was treated with natalizumab in the context of a clinical trial for Crohn's disease. After review of these PML-cases it was concluded that the risk of developing PML after use of natalizumab for approximately 18 months is 1 per 1.000.

When the first two cases of PML were announced on 28 February 2005 it was surprising to find out that little information had been published about the pharmacodynamic and -kinetic properties of natalizumab. An investigation followed to find out whether natalizumab interferes with immune surveillance of the CNS. The main conclusion of this investigation was that natalizumab reduces all lymphocytes subsets in the CSF. CD4+ T-cells and CD19+ B-cells were significally more affected than CD8+ T-cells and CD138+ plasma cells. However, the most interesting observation was that some pharmacological effects were still detectable even 6 months after discontinuation of natalizumab. It is currently believed that the almost immediate beneficial effect of natalizumab that has been observed in clinical trials probably has an effect on T-effector cells, CD4+ and CD8+, that migrate from the blood into the brain parenchyma. Prolonged use of natalizumab also has an effect on the migration and reconstitution of antigen presenting cells (APC's) in the cerebral perivascular space. After a certain number of doses this will lead to a significant reduction in the number of APC's and possibly eventually to complete depletion of these cells in the perivascular compartment. It is interesting to see that both MS patients that developed PML in the context of natalizumab received approximately the same number of doses (28 and 30).

14 months after the discontinuation of natalizumab therapy all cell counts in the peripheral blood and the CSF had returned to normal in the investigated cohort of MS patients. With regard to all clinical outcome (e.g. annual relapse rate, EDSS) and MRI measures the patients showed no difference while still on therapy and 14 months after they had discontinued. So clearly there was no rebound phenomenon, either immunologically, clinically or radiographic.

These results give rise to ideas about new treatment paradigms with natalizumab, especially with regard to the prolonged or indefinite use of these kind of drugs. One may wonder whether an induction therapy followed by treatment with other disease modifying therapies or followed by a treatment holiday may be an alternative. Clearly this would have to be demonstrated in the setting of a clinical trial.

Session 4. Target 3: Compartmentalization of inflammation in the CNS in multiple sclerosis. The brain as a special immunological compartment.

MEINL, Munich, Germany: Compartmentalization of inflammation in the central nervous system: a major mechanism driving progressive multiple sclerosis.

In the RRMS brain tissue active and inactive WM lesions and some cortical lesions with demyelination can be found. The brain of a progressive MS patient (PPMS or SPMS) is characterized by numerous inactive WM lesions in combination with diffuse cortical demyelination and injury in the NAWM (with perivascular infiltrates, micoglia activation and axonal spheroids). In 40% of SPMS patients follicle-like aggregates are found in the meninges (and not in RRMS or PPMS). Late stage lesions in chronic MS are enriched with B-lineage cells (Ig+).

Four lymphoid chemokines (CXCL13, CCL19, CCL21 and CXCL12) are known to be essential in the build-up of lymphoid tissue. These chemokines are also expressed in the CNS, but in a differential way. In normal CNS only chemokines CCL19 and CXCL12 are present which may play a role in immunosurveillance. In active MS lesions also CXCL13 is found. High level expression of CCL19 and CXCL12 is found both in active and chronic inactive MS lesions. These lymphoid chemokines recruit B-cells and other immune cells to the CNS.

B-cell activating factor of the TNF family (BAFF) contributes to the survival of B-cells in the CNS. It is a membrane bound molecule and has three receptors (TACI, BAFF-R and BCMA) that are essential for B-cell and plasma cell survival. BAFF may also play a role as co-stimulator for T-cells. BAFF is implicated in autoimmune pathology and has been established as a pathogenetic factor for SLE, RA, Sjögren's syndrome and Wegener‘s granulomatosis.

BAFF can be produced by astrocytes. In active MS lesions the expression of BAFF is as abundant as in lymphatic tissue (tonsils, adenoids). This all adds to the picture that BAFF production by brain resident cells might promote long-term survival of B cells and plasma cells in the brain of MS patients.

The inflamed CNS provides a survival niche for plasma cells. CXCL12 and BAFF are both important factors for the survival of plasma cells, as are other ligands and receptors that have been identified in MS lesions. When activated plasma cell also express a high level of α4b1 integrin. Plasma cells are very tough and difficult to attack therapeutically for they do not express CD20 and do not divide in the CNS. Their relevance for the disease process in not yet understood. Perhaps the inhibition of BAFF could be a therapeutic strategy. A BAFF-inhibitor will go into clinical trial in 2008.

Two recent articles described another interesting feature. In progressive MS the BBB shows subtle alterations outside of the perivascular cuffs. There is blood-brain leakage outside of active lesions in PPMS and SPMS. This can have two possible consequences. Influx of fibrinogen into the CNS will lead to microglia activation (via Mac-1). When the CNS can be accessed by antibodies this will contribute to increasing axonal injury. Recently antibodies against neurofascin have been showed to promote an axonal injury.

It is now increasingly becoming clear that all phases of MS are associated with inflammation. In RRMS the inflammation is focal. In SPMS there is global inflammation with diffuse WM injury (including axonal damage), cortical pathology, follicle-like aggregates in the meninges, a subtle leak of the BBB in the NAWM (with influx of fibrinogen and autoantibodies), combined with a CNS environment that fosters immune cells persistence (via BAFF and lymphoid chemokines CXCL12, CXCL13 and CCL19).

PLUCHINO, Milan, Italy: Can stem cells modify inflammation in the brain?

In the case of multifocal CNS demyelination one can think of three major strategies to achieve brain repair and immunoregulation. A first strategy would be to rescue as much endogenous WM oligodendrocytes around different types of lesions and try to achieve myelin repair. A second strategy involves the mobilisation of endogenous multipotent properties of stem cells from CNS germinal niches. This strategy is aimed at a shift in multigenetic pathways toward oligodendrogenesis. A third strategy would be the transplantation of exogenous (stem) cells able to accumulate in areas of CNS tissue damage in order to achieve myelin repair and local homeostasis.

Various animal studies have shown that neurostem cells show a remarkable therapeutic plasticity following systemic stem cell injection. They show effect on oligodendrocytes, inhibit astrogliosis and interfere with a number of immune mechanisms taking place inside and outside the CNS. These potential benefits of stem cells in EAE need to be carefully investigated and reviewed before a treatment strategy against MS can be developed. For this we have to take our time.

CLANET, Toulouse, France: Chemical substances that block macrophages and microglia in the brain.

Microglia are of mesenchymal origin and invade the brain during its early development. Population renewal occurs by intrinsic proliferation and by recruitment from the periphery. Microglia are critically involved with innate and adaptive immune system regulating inflammation and cell damage. There are two different populations of microglia: the perivascular and the parenchymal microglia. The perivascular microglia are directly derived from peripheral monocytes and macrophages. They have an APC phenotype and renew rapidly. The parenchymal microglia are the resident macrophages in the CNS. They are close to the neurons and axons and have a phagocytic phenotype. Their turn over from intrinsic proliferation is low. Microglia undergo morphological changes when they are activated.

Microglia can respond to internal signals (e.g. damaged or stressed cells) as well as external signals (e.g. pathogens). When activated they upregulate the expression of several molecules, some of which are quite specific (e.g. CD45, MHC II, CD40, B7 1 and 2, chemokines receptors CCR1, CCR2, CCR3). Microglia also produce innate cytokines (TNF alpha, IFN gamma), chemokines (MCP-1, MCP-2 MIP1-β), prostanoids (prostaglandins and thromboxanes) and nitric oxide (NO).

When resting, microglial cells constantly screen their environment. The term 'resting' does not properly reflect the constitutive surveillance activity of these cells, which would be better termed 'surveying' microglia. Their fine processes undergo continuous rebuilding to allow efficient scanning of their territory. Equipped with receptors for a plethora of molecules, they can immediately sense signs of disturbed structural and functional integrity.

Neurons deliver calming signals which keep microglia in this surveillance mode, indicating normal function.

In case of a small local neuronal damage microglia are activated through purinoreceptor ligands to produce neurotrophic factors which support neurons in restoring homeostasis. In case of a large insult to the CNS (infectious challenge or significant tissue injury) there is a strong reactive behaviour of microglia which may lead to impairment of neurons and glia. In fact, there is a graduation in the intensity of the reaction of microglia.

Macrophages and microglia play an important role in the development of neuroinflammatory agents. Neuroinflammation in EAE requires multiple steps of CD4+ T-cell activation in distinct CNS compartments. Naive CD4+ T-cells are activated by dendritic cells (DC's) in the periphery. Activated myelin specific T-cells penetrate the BBB and enter the perivascular space of the CNS. CD11c dendritic cells from the periphery populate the perivascular space and become perivascular macrophages. They present myelin antigen which reactivates the myelin specific T-cells. CD4+ T-cells release proinflammatory cytokines which activate resting parenchymal microglia which then release proinflammatory cytokines, cytotoxic mediators (NO) and chemokines which drive demyelination and secondary epitope spreading.

Microglia show features of a Janus face. They can exert a protective function in EAE and MS by secreting growth factors (BDNF, IL6, PDGF, GDNF). But depending on the signal and the microenvironment there is a graduated response of microglia leading to protective or detrimental reactions. There is no drug available that specifically targets the inhibition of microglia or macrophages. Animal studies have shown that depletion of peripheral macrophages abrogates the clinical symptoms of acute EAE. Also microglial paralysis has been shown to repress clinical EAE in transgenic mice. A number of ways have been considered to modulate microglia or macrophages, such as monoclonal antibodies, steroids, glatiramer acetate and statins. Especially minocycline has been indicated to show great promise in this respect. It is a powerful inhibitor of microglia activation and also provides some neuroprotection. When given at the time of disease induction experimental data showed a delay in neuropathological changes in EAE and attenuated the course of the disease. When given to a few MS patients reduced MRI activity was noticed. Unfortunately, minocyclin turned out to have a harmful effect on patients with ALS.

In conclusion it can be stated that microglia are an active sensor and a versatile effector in the normal and pathologic brain. Translational neurosciences is sometimes unsatisfactory owing to the limits of the transgenic mouse model. Modulation of macrophages/microglia is an exciting but risky new axis of therapy in MS.

ANTEL, Montreal, Canada: Impact on neural cells of immunomodulatory agents that cross the blood brain barrier

With regard to the effects of available and emerging immunotherapies against MS, we are moving into an era in which we have to consider how these agents may act on the CNS, whether they work in parallel and in which of the various compartments. These effects can be either direct or indirect. Indirect effects are effects in the CNS that are mediated via cells and products of the immune system that have been exposed systematically to immunomodulators . Examples in this category are glatiramer acetate (GA) and interferon beta. GA has a potential neuroprotective effect by acting on the immune system, either on lymphocytes or by modulating the monocytes microglia.

Direct effects are mediated by immunomodulators that directly access the CNS. Examples are chemotherapies, simvastatin, fingolimod (FTY720) and monoclonal antibodies. With regard to high dose chemotherapy with drugs that are able to access the CNS the effect on early changes of the brain volume is an important issue to consider.

There are various ways to describe the direct effects in the CNS. One is the effect on myelin homeostasis and myelin repair (remyelination). Most of the existing evidence suggests that remyelination within the CNS is caused by progenitor cells moving into lesions, rather than residing oligodendrocytes. In the oligodendrocyte lineage progenitors cells that express the ganglioside A2B5 on their surface give rise to astrocytes, neurons and oligodendrocytes.

Two examples of immunomodulators that affect CNS demyelination are statins and FTY720. When reviewing effects it is always a question whether the net myelin effect reflects a direct effect on oligodendrocytes and/or progenitor cells or an indirect effect of reduced immune mediated injury. Statins (HMG-CoA reductase inhibitors) have been shown to augment survival and differentiation of oligodendrocyte progenitor cells in animal model of MS. An in vitro survey of oligodendroglial responses to simvastatin showed initial enhanced process extension, followed by delayed process retraction and eventually delayed cell death, both in oligodendrocyte progenitor cells and mature oligodendrocytes. New data from in vivo studies with statins are expected in the near future.

FTY720 (sphingosine-1-phoshpate receptor agonists) attenuates relapsing-remitting experimental autoimmune encephalomyelitis in SJL mice. An in vitro survey has shown a direct supportive effect of FTY720 on the survival of oligodendrocyte progenitor cells. In terms of process dynamics an initial process retraction (signalled by the S1P5 system) was observed followed by process extension (signalled by the S1P1 system). How long this cycle continues needs to be further established.

It was already quite clear that MS presents a problem both on immunology and neurobiology. And now it has become apparent that also the MS therapies present us with a challenge in both these areas.

Session 5. Target 4: Brain tissue injury. Protection and recovery in the brain.

LASSMANN, Vienna, Austria: Mechanisms of inflammation induced tissue injury in multiple sclerosis

In the CNS there are 4 fundamentally different types of mechanisms of inflammation that all result in the same type of tissue injury, the inflammatory demyelinated plaque. This demyelination can be induced either by cytotoxic CD8⁺ (Tc1-) cells, CD4⁺ (Th1) cells, Th1 cells in combination with antibodies against myelin or by stimulating microglia through mechanism of innate immunity. In MS these types of mechanisms of inflammation are quite similar.

A first mechanism of tissue injury in MS is by induction of cytotoxic CD8⁺ T-cells. These cells need the expression of MHC class I antigens on their target cells in order to induce demyelination or even axonal injury. In acute MS lesions every cell type (neurons, oligodendrocytes, astrocytes and axons) can express these antigens. Whether this mechanism of tissue injury is restricted only to fulminant cases of acute MS is not known.

A second mechanism of tissue damage in MS is caused by specific antibodies. Only in combination with T-cell inflammation these antibodies are able to go through the BBB and attack their targets, either in cooperation with complement or with macrophages. An clear example of this mechanism is Devic's type of neuromyelitis optica, in which serum-antibodies (directed against aquaporin-4) recognize astrocytes at the glia limitans. This results in astrocyte specific destruction with astrocytic complement depositions, which is the key feature of the pathology.

Another example is the presence of auto-antibodies against neurofascin that can be present in MS patients. These antibodies do not cause much structural damage, but rather induce a functional blockade of nerve conduction by binding to the nodes of Ranvier.

A third mechanism of tissue injury in MS is mediated by innate immunity in which macrophages and microglia are involved. All kinds of macrophage and microglial products can damage myelin and axons, such as proteases, cytotoxic cytokines (e.g. TNF-alpha), excitotoxins and complement. Importantly, there is a differential vulnerability of CNS cells to macrophage toxins. The myelin/oligodendrocytes complex is more susceptible to these toxins than axons, neurons or astrocytes. Therefore, these lesions show extensive demyelination in combination with some axonal damage.

Additionally macrophages and microglia produce reactive oxygen and nitrogen radicals which cause mitochondrial impairment. This is a pathway that is of particular interest in MS pathology and pathogenesis of tissue injury. An attack of nitric oxide and oxygen radicals on the respiratory chain of mitochondria (specifically on complex IV) results in hypoxia in a background of inflammation. The production of these radicals by microglia is induced by fibrin deposits rather than T-cell inflammation as was suggested earlier. There is extensive evidence that fibrin is a potent activator of microglia (through TLR4- and Mac-1 activation) and stimulates radical production.

Also other TLR-ligand pathways of microglia activation have been described involving oxidised lipids and peptidoglycanes.

GONSETTE, Melsbroek, Belgium: Oxidative stress and excitotoxity a therapeutic issue in multiple sclerosis?

In MS pathogenesis microglial activation leads to two closely related pathogenic cascades: the inflammatory and the degenerative cascade. The inflammatory cascade leads to cellular (CD8, TNF-alpha) and humoral toxicity (antibody mediated + complement). The degenerative cascade leads to oxidative stress and excitotoxicity, which predominantly contributes to tissue damage in the secondary progressive phase of MS.

Oxygen is essential for aerobic life, but also a precursor of harmful Radical Oxygen Species (ROS). Oxidative stress reflects an imbalance between ROS production and natural anti-oxidant forces. Under pathological conditions elevated levels of ROS, produced during inflammatory processes (TNF-alpha, Interferon-gamma) or mitochondrial respiratory chain dysfunction, overwhelm natural anti-oxidant defences. The key effector molecule of oxidative stress is peroxynitrite (ONOO^-). Peroxynitrite is not a radical itself, but generates three free radicals ( carbonate, nitrogen dioxide and hydroxyl) that play a main role in oxidative stress. It has become generally accepted that peroxynitrite is a common effector toxic molecule in a large number of diseases including MS. In MS peroxynitrite is present in acute and chronic active lesions but is not detectable in inactive lesions. Potential peroxynitrite toxicity in MS causes membrane channel inhibition, calcium dysregulation, protein and lipid peroxidation and nitration, and mitochondrial dysfunction.

Glutamate is a principal Excitatory Amino Acid (EAA) for synaptic connections and neurological functions (cognition, movement). EAA overload leads to CNS dysfunction and lesions. Excitotoxicity reflects neuronal death caused by excessive exposure to EAA's (glutamate, aspartate). It results from a disruption of the glutamate-glutamine cycle in which astrocytes take up glutamate, convert it to glutamine (through ATP dependant mechanisms) and release glutamine in extracellular spaces. Neurons take up glutamine and convert it back to glutamate (by glutaminase). ATP-depletion in astrocytes results in increased intracellular glutamate and extracellular glutamate overload. Excitotoxicity and oxidative stress are closely interactive in neurodegeneration (by an auto-toxic loop). There are numerous experimental and biological data supporting the relationship between exitotoxicity and MS.

A number of potential therapeutic targets can be mentioned with regard to neurodegeneration. These include microglial inactivation, peroxynitrite catalysts and natural anti-oxidants (against oxidative stress), NMDA-antagonists, Ca⁺ antagonists and Na or K channel blockers (against excitotoxicity), mitochondrial protection and anti-apoptotic factors. Positive results obtained in the laboratory have not been translated successfully to the clinic. In fact clinical trials to protect the CNS against neurodegenerative processes have lead to inconclusive results: either modest effect (riluzole in MS), no effect in most trials and even negative effects (minocycline in ALS). It is clear that modulating a single specific pathway in a disease characterized by parallel mechanisms will likely yield a partial benefit only. It might well be that targeting proximal upstream events leading to the degenerative cascade turns out to be more rewarding (peroxynitrite catalysts, uric acid, tyrosine containing peptides). Clinical observations have shown that anti-inflammatory agents administered in the early phase of MS indirectly block early associated degenerative processes and delay the development of neurodegeneration. These are important observation for determining future therapeutic strategies.

KAPOOR, London, UK: Sodium channel blockers and neuroprotection using lamotrigine.

Membrane voltage gated channels have been a treatment target for a number of diseases (e.g. stroke, Parkinson's disease) for many years. Only recently it has been brought into connection with MS. It's only 7 to 10 years ago that neurodegeneration really became a key feature in MS. Until that time we had not been interested in axons all that much. Also the structure of axons (with its sodium, potassium and calcium channels and their distribution) hadn't been fully worked out. Moreover, recent insights in the complicated interaction between the function of the membrane and the metabolism of the axon have redirected our thinking. So there are a number of reasons why this field in MS is still comparably young, but we can learn from experience in other areas such as stroke.

In animal studies it was found that nitric oxide (NO) in pathological concentration causes a reversible blockade of axonal function. This finding was an initial indication that inflammation could interact with electrophysiology at an axonal level. But then NO also turned out to be a toxic compound for axons able to cause irreversible axonal degeneration. Another observation was that the metabolic strain put on axons by electrical activity could potentiate the effects of the inflammatory environment. So not only the electrical activity of the axons is important but also the metabolic strain that this activity puts on them in the toxicity of NO. The explanation of these phenomena lies with sodium channels. NO has many pathophysiological effects. One of them is the activation of sodium channels. When axons are placed in an inflammatory environment containing NO, sodium channels are activated and axonal polarisation is blocked. In MS axons are particularly vulnerable to this effect. Sodium channel expression is elevated after demyelination of axons. These axons are especially vulnerable to anything that acts on sodium channels. Another effect of NO that is of particular interest is that it inhibits mitochondrial respiration. So not only are these axons overloaded with sodium, but they are also unable to handle it. Sodium can only leave the axon through the reverse operation of the sodium/calcium exchanger, leading to calcium overload within the axon. This model occurring in MS is exactly the same as the situation in hypoxia. It immediately provides targets for therapeutic intervention such as inhibition of the sodium/calcium exchanger, calcium channels and especially sodium channels. Because of their overexpression, sodium channels will be key targets.

With regard to stroke the concept of neuroprotection has been spectacularly unsuccessful. One of the reasons is that neuroprotective drugs were put into clinical trials too early. Plausibility of mechanisms of action is actually not enough for drugs to be put into clinical trials. It was even the case that drugs were clinically tested without adequate preclinical testing of the human application (such as the ability of the drug to cross the BBB and determining relevant clinical endpoints). It can be stated that sodium channel blockade has satisfied most of the criteria for clinical testing. In vitro studies have confirmed that sodium channel blockade (with flecainide) does provide neuroprotection. Since flecainide proved difficult to apply in clinical practice (in view of cardiotoxicity), lamotrigine was further developed. The lamotrigine trial uses a standard randomised controlled, parallel-group design to test whether neuroprotection can be achieved in SPMS by partial blockade of voltage gated sodium channels. Decisions on the target population, primary outcome and sample size were dictated by the availability of data on rates of cerebral atrophy in two cohorts of patients, and by ethical and cost constraints. Present techniques for quantifying neuroprotection are imperfect, and the trial offers opportunities to evaluate a number of surrogate measures of tissue damage for use in future trials. The first results from the lamotrigine trial are expected by the end of 2008 or maybe beginning of 2009.

BLACK, West Haven, USA: Phenytoin protects central axons in EAE. Indications for a new treatment of multiple sclerosis?

The observation that axonal pathology occurs in MS plaques is not new. It was first described by Charcot in 1868. Only in the last decade the contribution of axonal degeneration to the acquisition of non-remitting disability in neuroinflammatory disorders (such as MS) has become fully appreciated. This has lead to the search of new therapies to protect axons. One of the proposed therapies involves the use of sodium channel blockers. There are four clinical trials with regard to sodium channel blockade. A first trial concerns the neuroprotection of lamotrigine in SPMS. Then there is a second trial concerning a combination of topiramate and interferon-beta-1a in patients with RRMS. A third trial involved the effect of phenytoin on axonal protection. This study was put on hold and eventually closed without further initiating it.

Phenytoin is a very old drug, first synthesized in 1908. In 1938 it was approved as an anti-epileptic drug. Phenytoin binds preferentially to inactivated sodium channels. It reduces peak currents, it is a voltage - and frequency dependant sodium channel blocker which has its efficacy in controlling seizures.

Some of the important observations in preclinical testing phase of phenytoin (using mice with EAE) included (in chronological order):

· Phenytoin improves clinical scores in MOG-EAE (short and long term)

· Axonal loss in MOG-EAE is attenuated with phenytoin treatment (short and long term)

· Phenytoin attenuates immune cell infiltrates (short and long term)

· Abrupt withdrawal of phenytoin acutely exacerbated EAE which results in acute high rate mortality.

· Withdrawal of phenytoin in not accompanied by axonal loss, thus implying that axonal degeneration is not a cause of mortality.

· Phenytoin withdrawal is accompanied by a large increase of inflammatory infiltrate (by factor 2-3).

· Phenytoin reduces vascular permeability. Withdrawal of phenytoin increases vascular permeability.

· Gradual withdrawal of phenytoin leads to gradual clinical worsening (with no effect on mortality rate).

Comparable observations were registered with the use of another sodium channel blocker, carbamazepine. Carbamazepine was shown to improve clinical scores in MOG-EAE and its withdrawal exacerbated EAE. However, this exacerbation was accompanied by a much lower mortality rate than was seen with phenytoin. Also the immune cell infiltrations after withdrawal were far less extensive than after the rapid withdrawal of phenytoin.

As to how phenytoin protects axons from degeneration and what mechanisms underlie exacerbation of EAE following phenytoin withdrawal, the answers are unclear. This is also the reason why initial plans for a clinical study were finally withdrawn. Sodium channels are not only expressed by neurons and axons, but also in a number of other cells such as immune cells, astrocytes and endothelial cells.

In conclusion phenytoin provides long-term improvement of clinical status, prevents axon loss and ameliorates immune infiltrates in MOG-EAE mice. Sudden withdrawal of phenytoin leads to acute exacerbation of MOG-EAE and extensive immune infiltrates. It is likely that phenytoin has multiple sites of action, which should be taken into account when considered as a treatment for MS.

FRANKLIN, Cambridge, UK: Protection and restauration of the brain by stem cells.

There is general consensus that axons need to be protected in MS. In acute MS lesions rapid and extensive axonal injury is associated with inflammatory response. There is also slow progressive axonal loss in chronic demyelination which is caused by chronic low grade inflammation. An alternatively mechanism for axonal loss in chronic demyelination is a loss of myelin trophic support by which the axons become more vulnerable to degeneration. Remyelination is a process by which demyelinated axons are restored. This is why remyelinating therapies are viewed as an important component of the MS treatment armory. For reasons that are unclear and undoubtedly multiple, remyelination by oligodendrocytes fails in some circumstances. As a result axons become chronically demyelinated and vulnerable to degeneration. This can be prevented by making sure that remyelination takes places so that axons can functionally recover.

There are a number of strategies available for promoting remyelination in demyelinated injury. One of the strategies that have been investigated for many years is to transplant cells with a myelinating capacity in areas of demyelination. A variety of different cell types (e.g. neural stem cells, precursor cells, differentiated cell types ) have been used, injected in a variety of routes, either directly of indirectly. Another strategy is by stimulating repair by endogenous stem/precursor cells that normally remyelinate chronically demyelated lesions. These cells are abundant and widespread in the CNS. This strategy involves the targeting of these cells to do their work more efficiently than they do in some cases of MS.

The oligodendrocyte precursor cells (OPC's) are the stem/precursor cells that are fundamentally important in the process of remyelination. When inducing primary demyelination by use of toxins in an experimental rodent model, there is a robust and rapid response of the endogenous precursor cells. These cells are activated in response to initial activation of microglia and astrocytes and the factors that they produce. When activated these precursor cells switch on a specific set of transcriptional regulators (such as Nkx2.2, Olig2 and Sox2). In response to nitrogen generated by the innate inflammatory process they proliferate and are recruited into the lesion very efficiently. These precursor cells differentiate into new oligodendrocytes that repair the focal demyelinated lesions by remyelination. And therefore these precursor cells are important cells that can be targeted therapeutically to make remyelination more efficient.

OPC's and stem cells have several features in common. Therefore OPC's could be regarded as adult neural stem cells. Like stem cells OPC's are able to renew themselves, express stem cell markers, are multipotent (meaning that they are not committed to differentiate into one single phenotype) and contain high levels of telomerase. It is yet uncertain whether OPC's do also undergo replicative senescence like stem cells do. However, unlike stem cells OPC's do not divide asymmetrically or slowly after injury.

In the effort to make stem cells more effective in remyelination, the recruitment of precursor cells does not present a major problem. The vulnerability of this strategy lies with the differentiation of precursor cells to become oligodendrocytes. It is believed that there are specific factors within chronically demyelinated lesions that inhibit differentiation of precursor cells. A number of these factors (such as PSA-NCAM, Hyaluronan, Lingo-1 and Notch-jagged) have been advocated in this respect. However, functional evidence that these factors are indeed inhibitors of remyelination is somewhat absent.

Another hypothesis for the lack of differentiation of precursor cells is the absence of differentiation signals. A number of growth factors are associated with the induction of differentiation of precursor cells, IGF-I probably being the most important. A variety of animal experiments have made it clear that there is an abundance of redundant components in the signalling environment of remyelination. Changing just one of these components is insufficient to make the remyelinating process work more efficiently. Neither inhibiting factors (such as Notch-jagged) nor signalling factors (such as IGF-I) solely determine the efficiency of the process.

The high level of redundancy in the signalling environment of regeneration processes (such as remyelination) has profound implications for the design of therapeutical targets for promoting remyelination. Finding non-redundant components on which the process depends is a challenge. One identified non-redundant mediator of remyelination is transcription factor Olig-1. Its function, however, depends on a matrix of other (either upregulating or downregulating) transcription factors. When trying to identify targets for therapeutically promoted remyelination, given the high level of redundancy in the regenerative process, it is probably wise to move away from thinking in terms of single targets and single culprits for remyelination failure. Instead a more subtle manipulation of a more complex matrix of factors acting in concert should be considered.

Session 6. Target 5: How to measure treatment effects in multiple sclerosis? Are the ends clear enough to justify the means?

MILLER, Haifa, Israel: Translation towards personalized medicine in multiple sclerosis.

In MS there is no clear pathway for prevention or cure. Therefore current therapeutic interventions are aimed at controlling the progression of the disease by immunotherapy, neuroprotection, immunosuppression, neurorepair (stem cells), symptomatic treatment (to improve quality of life) and/or combination treatments (cocktails). In recent years there has been a shift from treating the disease (by treating disease subtypes and specific patient populations) towards treating patients (personalized medicine). One example is genomic profiling of interpopulation diversity in order to prioritize candidate-genes for autoimmunity. Relevant choices with regard to the medical treatment of MS are what medication to use, how much (dose), by what route (i.m., i.v., s.c., p.o.), with what (cross reactivity, combinations), when to start, how long to use and when to stop.

Another relevant issue could also be at what time of day treatment should best be administered. The neuro-endocrine immune network has a biological clock with diurnal variations. We know that cortisol levels normally rise and fall during the day, repeating a 24-hour cycle (diurnal variation). The highest cortisol levels are between 6-8 a.m. and the lowest levels are around midnight. Little is known about the diurnal variation of cytokines and their receptors or of soluble adhesion molecules in healthy individuals or patients. In MS the hypothalamo-pituitary-adrenal axis (by which cortisol levels are controlled) is dysregulated. The basic secretion rate of cortisol in MS patients lies within the normal range. However, reactivity of the adrenal cortex to stress and stimulation by synthetic ACTH is markedly reduced. The function of the biological clock and the circadian variations in MS are yet to be elucidated. There are indications (from a small sized study with RRMS patients receiving IVMP treatment) that night time treatment (at a time when cortisol levels are down and T-cell reactivity is high) provides better results than daytime treatment, both in terms of tolerability and efficacy. These findings are supported by other immunological data.

Our judgment of response to MS treatment is based largely on clinical parameters. Therefore the clinical definition of response is very important. Various definitions of responsiveness to MS medications exist. Instead, it seems easier to define non-responsiveness to MS treatment. A distinction should be made between primary and secondary non-responsiveness. Primary non-responsive patients are those that are drug resistant and do not respond to treatment from the beginning. Drug resistance should be distinguished from drug intolerability or drug allergy. In secondary non-responsiveness patients initially seem to respond well to treatment, but after a few months or years respond less. This may be due to change in disease activity or subtype (for example from inflammatory to more neurodegenerative or from cerebral to spinal activity). Non-responsiveness may be also caused by development of neutralizing antibodies of drug-drug interactions or metabolic changes (e.g. hypo- or hyperthyroidism).

A key question is whether early responsiveness predicts long-term responsiveness. Biomarkers may help to provide the answers to this question. A lot of research is spent on which biomarkers to use and their relevance in view of the various processes (inflammation, demyelination, remyelination/neural repair, neurodegeneration) of the disease that simultaneously take place in RRMS. At which molecular level (T-cell activation, DNA, RNA, intracellular proteins, cell surface proteins or extracellular proteins) should biomarkers be developed and what is their relevance with regard to clinical decision making? Oligoclonal bands have been used as biomarkers for many years although their relevance to the pathogenesis of MS remains unclear. This raises the question whether we should aim at sampling for antibodies in MS patients. Not only oligoclonal bands are of interest, but maybe also anti-myelin (MOG), antiviral and remyelinating antibodies or auto-antibodies. The role of antibodies and their relevance as biomarkers or in choosing treatment is a major and complex issue which remains to be solved.

We know that response to treatment is multifactorial. It depends on (multi)genetic, environmental and epi-genetic factors. Pharmacogenetics is the study of variability of drug response attributed to hereditary factors in different populations. Part of pharmacogenetics is to find different genetic characteristics of responders, poor responders and adverse responders. It opens the possibility to move from a trial and error treatment approach towards a more informed way of predicting response to treatment, both in terms of efficacy and safety.

In conclusion optimization of MS treatment requires adjustment to disease subtype/activity, chrono-biological principles, specific populations, personal indicators (clinical, MRI and biomarkers of response) and pharmacogenetic markers. The development of a personalized medicine approach requires an integration of all these aspects at an individual level. Personalized medicine is an exciting and promising direction to move towards to.

CZAPLINSKI, Basel, Switzerland: Reconsidering clinical outcomes: relapses, impairment, disability and beyond.

The overall quality of clinical trials in MS is increasing steadily. This is due to the growing experience in the area, the increasing awareness of quality standards within the MS community and the more stringent requirements of regulatory authorities for approval of new treatments. Each successful clinical trial has provided additional information that could be incorporated into the design of subsequent studies to improve their quality. New studies present a further series of challenges related mainly to the choice of appropriate outcome measures, the use of placebo and the choice of the comparator, the extent of treatment exposure and the assessment of side-effects.

Outcome measures for clinical trials in MS are either relapse or disability (impairment, handicap) related. Several definitions of relapse exist. One that is used is a situation of new or worsening neurological symptoms confirmed by objective findings on neurological examination. Duration of symptoms has to be longer than 48 hours and criteria for neurological changes are defined as an increase in EDSS >0.5 points and/or >1 point increase in FS. A first issue in registering relapse rates is the highly positive placebo effect (up to 50%) found in pivotal and randomized trials. Is this a genuine placebo effect, or is it perhaps dependant on blinding efficiency, regression to the mean or comprehensive care? Another issue is that relapse rates depend on the frequency of clinical examinations. A higher frequency of clinical visits increases the probability to detect clinical relapses and consequently results in higher relapse rates. Relapse rate can be expressed in annualized relapse rate, number of relapse-free patients or time to first relapse under treatment. Annualized relapse rate is not suitable in trials with an expected high dropout rate. Patients that discontinue the trial after 1 or 2 relapses may bias the results by increasing the annualized relapse rate. Time to first relapse under treatment is a robust parameter, even in trials with high dropout rates. However, this parameter may favour drugs that are effective shortly after the beginning of treatment. It does not measure effects occurring later on in the course of treatment. Inaccuracy in measuring relapses can be minimized by clear definition of a relapse, if possible, in combination with objective confirmation. Patients must be asked whether they experienced some relapse-suspected signs between two clinical visits. Observation time should be long enough. Treatment effect on relapse rate should be measurable after one year of trial duration.

Disability is an activity limitation, defined as any restriction or lack of ability resulting from an impairment to perform an activity in the manner or within the range considered normal for human beings. The ideal outcome measure for disability is multidimensional, applicable across the range of disease severity, easy to use and reliable. Also it should have some predictive value and its individual components should be sensitive to change over time.

The Expanded Disability Status Scale (EDSS) is the most frequently used scale for rating disability in MS and is accepted by the FDA and EMEA. In its lower ranges (0 – 5.5) EDSS is defined by functional systems, in the middle ranges (4 – 7.5) it is defined by walking distance and in the higher ranges (7 – 9.5) it is defined by dependence on help. Advantages of the EDSS scale is that it is the longest existing (for more than 40 years) and most widely used rating scale. Most MS specialists are familiar with the scale, it is easy to use and it allows a simple comparison on a scale from 0 to 10. Disadvantages of the EDSS scale are its poor inter- and intra-rater reliability, and its bimodal distribution of scores. EDSS is based on an ordinal scale (which means that changes between numeric steps are not all the same) and it relies heavily upon ambulation. Also there is a poor assessment of upper limb function, it is insensitive to cognitive decline and the reproducibility in the lower ranges of the scale is low. In Clinical trials the reliability of EDSS can be increased by letting the same physician evaluate individual patients throughout the trial, measure walking distance during the exam, standardized schedule and time of assessments, standardized documentation of the neurological examination (with the Neurostatus form) and regular EDSS training and certification.

The Multiple Sclerosis Functional Composite (MSFC) is a newer scale, designed by the US National MS Society and driven by an extensive evaluation of previous MS trials. It has been more sensitive than EDSS in some clinical trials and outcome studies. The MSFC changes have also been found to reflect the severity of the disease, as perceived by patients, through the quality-of-life questionnaire (QoL-54). The MSFC score captures three different dimensions: ambulation (time to walk 25 feet), arm function (Nine-Hole-Peg test) and cognition (PASAT-test).

MSFC is basically a composite score that indicates how many standard deviations a subject is above or below the mean of the standard population on arm, leg and cognitive functions. Advantages of the MSCF rating scale is that it contains leg, arm and cognitive scores and it changes more quickly than clinical ratings (such as the EDSS score). Also it is a continuous scale which may be more informative than an ordinal scale and facilitates analysis of longitudinal data. Disadvantages of MSCF are that it is based upon a patient population analyzed in the original study in 1999. Also the effect of learning in the PASAT score is still uncertain and it does not involve a visual system. MSFC captures three domains of neurological function, which are affected in nearly all patients in later stages of MS (ambulation, dexterity of upper extremities and attention/short-term memory) but are less frequently affected in patients with earlier-stage disease, where symptoms such as visual or sensory deficits are frequently found in isolation. Therefore MSFC is not equally suitable for all MS subpopulations. The value of this scale may be improved by adding a fourth parameter, but this may also impact the power of MSFC used as a composite score.

The Multiple Sclerosis Severity Score (MSSS) corrects EDSS for the duration of the disease by using an arithmetically simple method to compare an individual's EDSS with the distribution of EDSS in cases having MS for the same length of time. Effectively the MSSS assigns to each EDSS its median decile score within this distribution. A MSSS of 5.0 means that the disease is progressing at a median rate. A patient whose MSSS is 9.0 is a fast progressor, progressing faster than 90% of patients. A patient whose MSSS is 1.0 is a slow progressor, progressing slower than 90% of patients. As a next step MSSS was introduced as a new severity-based classification system for multiple sclerosis (MS) similar to the staging system used for cancer, ranging from benign MS (MSSS score of 0.45 or lower) to malignant MS (MSSS score of 9.59 or higher). In between categories of moderate, intermediate, advanced, accelerated and aggressive MS. This score may be useful in clinical trials to define subpopulations of interest (e.g. exclude benign and malignant courses) or even as outcome measurement.

In a given trial multiple outcome measures should be determined. It has been a consistent finding of the successful positive clinical trials that most, if not all, of the outcome measures respond in a coherent way. This considerably strengthens the relevance of the findings. It is important to choose the most appropriate primary outcome measure for each individual trial and to use a sample sufficiently large to provide the power necessary to demonstrate effect. The outcome measure may depend on the presumed mechanism of action of treatment under consideration and its anticipated clinical activity.

TINTORé, Barcelona, Spain: New MRI measurements for the treatment effects on relapses and progression.

MS is considered to be a two stages disease. The first stage in which inflammatory events are crucial represented clinically by relapses and the second stage in which degenerative phenomena are the rule represented clinically mainly by progression.

After a CIS, at the initial phase of the disease, a baseline MRI can be useful in predicting the risk of conversion to clinically definite MS (CDMS) as well as the development of disability, to monitor the effect of the treatment and to some extend to predict response to treatment. Not only is MRI able to tell us the proportion of patients who are going to convert to CDMS but also the amount of time for this conversion to take place. Patients with more lesions at baseline MRI (with 3-4 Barkhof criteria) are going to convert in a very high proportion but also very rapidly. Oppositely, patients with no lesions are going to convert in a very low proportion and if they do so they are going to convert only in the long term. In the Barcelona CIS cohort a higher risk for developing disability was demonstrated in patients with a baseline MRI with 3-4 Barkhof criteria. There was a moderate correlation between EDSS 5 years after onset of the disease and the number of Barkhof criteria at baseline.

MRI is quite useful in clinical trials to monitor the effect of the treatment and also in determining which patients are going to respond better to treatment. A study performed in Barcelona showed that the development of new T2 lesions during the first year after a CIS may be a marker for bad response to IFN treatment in early MS.

In RRMS MRI is of little use in predicting relapses or development of disability. However, it remains useful in monitoring treatment effect and in predicting response to treatment. Clinical predictors for disability increase and/or relapses usually take some time to be evaluated. MRI predictor can be very helpful in that respect.

A study performed in Barcelona showed that MRI changes (> 2 active lesions) occurred during the first year of IFN treatment may have a prognostic value for identifying patients with a confirmed increase of disability after two years of therapy.

In SPMS, the more degenerative phase of MS, it is more difficult to establish the role of MRI. Relapses and new T2 lesions are no longer helpful as markers for disease progression. When reaching an EDSS of 4, new T2 lesions show a plateau effect for higher EDSS scores. Therefore new markers have to be developed to be able to follow the course of the disease in its later stages. Several markers have been investigated, but available data are too limited to make any recommendations. The use of magnetization transfer looks promising, although it is not possible to draw any conclusions. A current technique used in clinical trials for monitoring neurodegeneration is atrophy measures. Spectroscopy is a challenging technique for the future, but is not usable across centres studies.

After a CIS, MRI can be useful for predicting the risk of conversion to CDMS, to predict the development of disability, to monitor treatment effect and to some extend to predict response to treatment

However, we have much more difficulties in finding a reliable surrogate MRI marker in the more degenerative phases of the disease. Many efforts are made in the context of the MAGNIMS group to standardize new MRI techniques. It is only a matter of time before better MRI markers will become available.

LEOCANI, Milan, Italy: Neurophysiology in the measurements of treatment effects

Functional deficits in MS pathology rely on segmental demyelination and axonal degeneration. These pathological substrates lead to functional effects on evoked potentials (EP's). Demyelination leads to conduction slowing (reflected by latency of EP's), temporal dispersion, increased refractory period and conduction block. Axonal degeneration, increased refractory period and conduction block result in amplitude abnormalities in EP's. Conduction block almost invariably occurs if the length of the demyelinated area exceeds 5 mm, although the lesion size by itself does not predict the amount of functional loss. Conduction block may also be caused by soluble mediators of inflammation (NO) especially if axons are demyelinated. In those cases conduction blocks may be transient and reversible.

Thus, it is possible to collect information about the functional impairment of axons and the type of abnormality. Normal somatosensory EP's exist in the early stages of MS, when the CNS pathways are not yet affected by lesions. Latency occurs when components of the CNS pathways are affected, resulting in a delay of conduction. In this stage the patient may still be asymptomatic when the delay is not causing a functional deficit of which the patient is already aware. Functional loss starts to occur in case of conduction block or axonal loss, leading to loss of components. The amplitude of EP's can provide information about the number of axons that are still available for conduction. In the worst cases encountered in seriously affected patients with longer disease courses, when both cervical and cortical components are lost, a ceiling effect is reached for the neurophysiologist. In these patients the worsening of the disease can no longer be followed when the monitored components have already disappeared.

Several tools are available for the neurophysiological assessment of regional damage in MS. Among them are EP's, such as visual (VEP), somatosensory (SEP) and motor EP (MEP). SEP and MEP provide useful information because they test a longer pathway (i.c. the spinal cord). The VEP is a test for the function of the optic nerve which is usually involved in MS. VEP has a high sensitivity in MS patients with various disease courses (RRMS, SPMS and even PPMS). This is why in the new diagnostic criteria for MS VEP's are considered useful in cases of equivocal or negative MRI to determine the diagnosis MS.

EP's are quite useful in testing function of eloquent pathways in patients. There is a positive correlation between clinical and neurophysiological findings, as is seen between vibration sense and somatosensory EP, between motor central conduction time and hand dexterity and between VEP amplitude and lesion length in acute optic neuritis. The degree of severity of abnormality can be expressed in EP scores ranging from 0-3 (0 = normal; 1 = abnormal latency or amplitude; 2 = abnormal latency and amplitude of a major component; 3 = absence of a major component). By quantifying the severity of abnormality of the involved pathways it is possible to follow MS patients over time. Global EP, a combined score of all EP's, has a positive correlation with EDSS. Basal EP's can even predict EDSS changes in 4 years time. A high severity of involvement of SEP, VEP and MEP at the initial stage of the disease (or even subclinically) predicts a poor outcome on the development of disability (as expressed in EDSS). Basal MEP can predict EDSS after 6 months. With EP's it is also possible to focus on a selected pathway and follow it over time. After acute optic neuritis, for instance, a reduction of the latency potential can be observed during a 3 year follow-up period in the affected eye. It is also possible to observe any slow worsening of the unaffected eye. So with this method subtle abnormalities can be followed over a prolonged period of time, even in the absence of a clinical correlate. And also this method is capable to follow very slowly occurring changes over a period of 3 years. This provides us with a tool with which to assess any improvement over time as a result of remyelination. This could be useful for testing neuroprotective drugs in the future, providing additional information that can not be obtained by any other method.

EP's have been studied to assess the effects of therapy. It has been found that electrical assessment can provide a very early indication of the effects of therapy (by improvement of conduction) and also of the mechanism of recovery. Neurophysiological methods can also be used to assess acute effects of therapy. Cognitive functions can be affected by inflammation. By measuring cognitive event-related potentials it is possible to assess short term changes of anti-inflammatory treatment (e.g. steroids) on the cognitive function. Another possibility of this method is to predict the effect of therapy on selected functions. It has been shown that cognitive event-related potentials can be predictive for the response to modafinil treatment in patients affected by fatigue. Only responders to modafinil showed a good amplitude of the event-related potential.

EP's can be used as an outcome measure in clinical trials. They are very sensitive, especially in the optic nerve and spinal cord, and are well correlated with measures of disability (EDSS and FS). They are sensitive to subclinical changes and are a direct measure of function of sensorimotor pathways.

However, there are also a number of problems in applying these methods in clinical trials. There is a high variability among laboratories, sensitivity to disease activity may be low (only changes in investigated pathways are detected), in MS patients they could have an increased test-retest variability, and there are ‘floor' and ‘ceiling' effects (i.e. scarce proportionality between EP findings and disease progression in very early and late stages of the disease).

GIOVANNONI, London, UK: The role of biomarkers in treatment effects

A biological marker (biomarker) is a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.

Biomarkers are used as surrogate endpoints to clinical endpoints, and are expected to predict clinical benefit (or harm, or lack of benefit or harm) based on epidemiologic, therapeutic, pathophysiologic or other scientific evidence.

Apart from locally synthesized oligoclonal bands (OCB) used as part of the diagnosis workup of MS, biomarkers have been given less emphasis in the recent diagnostic criteria. There aren't any real biomarkers (except for two) that have made it to mainstream clinical practice. In 2004, around the time of the publication of a review article about the development of body fluid biomarkers in MS (by Bielekova and Martin), a pan-European consortium for CSF Biomarkers Research in MS (http://www.bioms.eu/) was formed to standardize, validate and improve biomarker research in MS.

An example of biomarker discovery is a study in which a surrogate endpoint is searched to determine clinical response and non-response to IFNb. The question asked is whether we can use some short term gene induced products (like Neopterin, Mx, IL10) or whether we should use more long term altered cell populations (like dendritic cells, NK-cells, B-cells, Th1, Th2, Th17 or Treg cells) to measure therapeutic effects. Examined biomarkers included cell populations (by cytometry), soluble proteins (proteomics, immune assays), small molecules (mass spectometry), RNA (not performed yet) and DNA (not performed yet). In this study a high throughput microvolume laser scanning cytometry system, called SurroScan, was used for cellular phenotyping. Study samples were tested on patterns of biomarkers for IFNb activity and for MS response and non-response to IFNb. MxA was used as a positive control for IFNb activity.

One of the most consistent things that come out of the literature regarding MS patients as a whole is that they have a fundamental deficit of NK cells. It remains a question whether this is an associated factor of whether it is linked to the pathogenesis of the disease. NK cells could be a good biomarker for MS therapy if a link could be made between successful MS therapy and the reversible effect on NK-cell deficit. However, this link was not found when applying chronic IFNb treatment. Several studies have made it clear that MS biomarkers cannot be analysed in isolation. Instead, we have to start analysing biomarkers in networks. There must be a whole line of gene networks that regulate NK cells. These need to be analysed in parallel with cell surface antigens.

Biomarkers are alternatively used in monitoring drug washout (e.g. clearance of natalizumab), prediction of responsiveness (e.g. neurofilament levels as biomarker for neuroprotection), or lack of efficacy (e.g. neutralising antibodies (Nabs) as a well documented predictive marker for the long term efficacy of IFNb therapy). That current knowledge about biomarkers (e.g. Nabs) is not yet fully adopted into clinical practice across the world is due to medicine (and the use of biomarkers) being a social science. This means that diffusion of innovation takes time to go from early to wide adoption.

In conclusion, widely accepted body fluid biomarkers in MS are locally synthesized OCB (as diagnostic marker), anti-natalizumab antibodies (as prognostic marker for infusion reactions and loss of efficacy) and neutralizing anti-IFNb antibodies. The BioMS consortium is encouraging the scientific process to develop new biomarkers. Lessons can be learnt from other fields, such as oncology and cardiology. Discovery based platforms are very exciting, the data are difficult to interpret and we need to move away from a single marker towards gene network analysis. The way forward is by hypothesis driven research.

LIST OF MEDICAL TERMS

Apoptosis A form of programmed cell death, involving a series of biochemical events leading to a characteristic cell morphology and death

Astrocyte Supporting cell in the central nervous system

Axon Neuron fibre that transmits an electrical signal to an endpoint (e.g. organ)

Chemokines Family of small cytokines or proteins secreted by cells (chemotactic cytokines)

Cytokines Group of proteins and peptides that are signalling compounds produced by human cells to communicate with one another

Glia Supportive tissue of the brain

- astroglia: astrocytes

- oligodendroglia: oligodendrocytes

- microglia: cells of the immune system of the brain

Intrathecal In the corticospinal fluid compartment of the spinal cord

Myeline (sheath) Insulation tissue around neurons to allow conductance

Neural precursor cells Cells that develop into neurons

Oligodendrocyte Cell that myelinates neurons

LIST OF ABBREVEATIONS

ACTH Adrenocorticotropic hormone

ALS Amyotrophic lateral sclerosis

APC Antigen presenting cell

ATP Adenosine 5'-triphosphate

BAFF B-cell activating factor of the TNF-family

BBB Blood brain barrier

CDMS Clinically definite MS

CIS Clinically isolated syndrome

CNS Central nervous system

CSF Cerebrospinal fluid

DC Dendritic cell

DNA Deoxyribonucleic acid

EAA Excitatory amino acid

EAE Experimental allergic autoimmune encephalomyelitis

EDSS Expanded Disability Status Scale

EMEA European Medicines Evaluation Agency

EP Evoked potential

FDA Food and Drug Administration

GA Glatiramer acetate

Gd / Gad Gadolinium

GI Gastro-intestinal

Ig Immunoglobulin

IL Interleukin

IM Intramuscular

IFNb Interferon-beta

IFNg Interferon-gamma

IP Intraperitoneal

IV Intravenous(ly)

IVMP methylprednisolone IV

MBP Myelin Basic Protein

MEP Motor evoked potential

MS Multiple Sclerosis

MSFC MS Functional Composite

MSSS MS Severity Score

NAWM Normal appearing white matter

NO Nitric oxide

OCB Oligoclonal bands

OPC Oligodendrocyte progenitor/precursor cell

PBMC Peripheral blood mononuclear cell

PLP Proteolipid Protein

PML Progressive multifocal leukoencophalopathy

PPMS Primary progressive MS

QoL Quality of Life

RNA Ribonucleic acid

ROS Radical oxygen species

RRMS Relapsing remitting MS

SC Subcutaneous

SEP Somatosensory evoked potential

SPMS Secondary progressive MS

TCR T-cell receptor

TGF Transforming growth factor

TNF Tumour necrosis factor

Treg Regulatory T-cell

VEP Visual evoked potential