European Charcot Foundation Symposium, 12 – 14 November 2009, Lisbon, Portugal.

"A new Treatment Era in Multiple Sclerosis: Options, Challenges and Risks”

Summary of lectures

XV European Charcot Foundation Lecture

CLANET, Toulouse, France: Trends in treatment strategies

In the latest issue of Nature about drug discovery an editorial mentions that five new oral MS treatments are expected in coming years. This means that we are now entering new period of MS treatment.

Historical view

Looking back at the past two decades of MS research, it is clear that we have accumulated various pieces of information to help us understand the MS puzzle. But still we do not know exactly what MS actually is. So far we observe MS as an heterogeneous disease, which is the result of complex interactions between inflammation and degeneration. We know that the immune system plays a major role in the pathophysiology of MS. There are two stages of this disease. The first is related to the relapsing-remitting course and is driven by inflammation. During this phase a major role is played by the adaptive immunity (antigen specific T-cells, B-cells). During the disease there is a shift in the spreading of the immune reactions. This results in the second stage, in which the disease becomes progressive. During this phase the innate immunity (with microglia, macrophages, monocytes, dendritic cells and various cytotoxic compounds) is playing a major role. Only recently it has been proven that MS is not only a disease with focal white matter lesions, but also with cortical lesions and abnormalities in the normal appearing grey matter during the second phase of the disease. Finally, there is a complex balance between inflammation with adaptive immunity in the relapsing-remitting stage of the disease and degeneration, which probably starts from the beginning of the disease but is amplified in the secondary progressive stage of the disease, and which is dominated by the innate immunity. We suppose now that the inflammation is the most important factor and causes degeneration, but it is also possible that neurodegeneration and inflammation play separate roles in the evolution of the disease.

In animal research we do not yet have the ideal MS animal model. Our EAE model is not sufficient for the development of new treatments. EAE is a poor predictor for success in MS. The complex heterogeneity of MS and EAE differs considerably. With new transgenic mice we are now approaching the model that is more related to MS. In the past two decades there have been many examples of treatments that have showed effectiveness in EAE but not in MS patients. Examples are anti-TNFa, IL-10, anti CD40 (IDEC-131), APL NBI 5788, ustekinumab and IL-23. Only a minority of treatments designed from EAE have been proven successful in MS. And so preclinical research is merely exploratory.

Early immunomodulation in MS started with a publication of a pivotal INFb-1b RCT in Neurology in April 1993. In the following five years INFb and GA were licensed for the treatment of RRMS patients. In 1995 the ETOMS trial (INFb-1a 22 μg/week versus placebo) was designed together with the ECF Board. In 2000 CHAMPS (Jacobs, 2000) was published and INFb-1a IM was licensed for CIS treatment in the next 2 years. New criteria for the early treatment of MS were published in 2001 by McDonald. In the BENEFIT trial (INFb-1b versus placebo) all the patients were offered active treatment after two years because of positive findings in delay of disability progression (based on EDSS). However, after 5 years follow-up the differences in disability progression (EDSS) had almost disappeared. And so it turned out that it would take 25 RRMS patients (NNT) and five years of active treatment in order to prevent disability progression by 1 EDSS point in one patient. On the other hand there have been several observational studies that showed a significant effect also after long-term treatment, but for several reasons it is difficult to draw conclusions from these findings.

It is difficult to compare the efficacy of treatments based on outcomes from different RCT's. Actually, it is virtually impossible to compare different drugs unless this is based on head-to-head comparison. The natalizumab case is a clear example that RCT's are of limited value for the identification of severe adverse events of low incidence (i.c. PML).

Current strategies

The current strategy for MS treatment involves two concepts: escalation versus induction therapy. There is major consensus among experts in the US and Europe in favour of escalation therapy which consist of a build-up in aggression of the approach. The first line therapy is INFb or GA. Second line therapy is mitoxantrone, cyclophosphamide or natalizumab. Third line therapy is a combination therapy and fourth line therapy is bone marrow transplantation.

The induction concept is based on the assumption that the early phases of MS are important, particularly the immune aggression at the beginning of the disease. This concept is designed to reset (or modify) the immune system by treating patients quickly and more aggressively from the beginning. And so it would be acceptable to treat patients in the first line therapy with mitoxantrone, cyclophosphamide or natalizumab followed by a maintaining therapy with IFN or GA (second line therapy). Again, third line therapy is a combination therapy and fourth line therapy is bone marrow transplantation. This induction concept still needs confirmation from clinical research. In the current phase III alemtuzumab trial (vs. INFb) this concept is being investigated after the phase II trial had shown positive results in reduction of disability progression (NEJM, 2008). Currently, there is consensus (Boster, 2008) that induction therapy should be started only in selected MS patients who have a very active disease.

New treatments and trends in strategies

There are quite a few new treatments in the pipeline, some of which may already become available within a year. Laquinimod, fumarate, fingolimod, teriflunomide, cladribine, alemtuzumab and rituximab are all in phase III of clinical development. In the escalation strategy laquinimod and fumarate (BG-12) are potentially first line treatments, while cladribine, fingolimod, teriflunomide and alemtuzumab may qualify as second line therapies. In the induction strategy cladribine, fingolimod, alemtuzumab and rituximab are potentially first line treatments, while laquinimod and teriflunomide could qualify as second line therapies.

Early market access vs. risk-benefit assessment

With regard to new treatments, neurologists are expecting to be able to slow down disease progression with efficient treatments: i.e. low risks, better efficacy and at low costs. Early market access of new drugs increases the uncertainty about the risk-benefit ratio, especially the uncertainty of safety. RCT's are superior to evaluate the benefits, but not the risks. Reasons for that are that RCT's are too brief, narrow and simple and the age variation and patient numbers are limited. Adverse drug reactions can therefore be missed in RCT's. This puts regulators in a dilemma whether to shorten or delay market access. Arguments in favour of shortening time to market access are that it creates favourable conditions for innovation (industry), that potentially life-saving treatments become available as soon as possible (patient groups) and that unmet medical needs are to be attended. Arguments in favour of delaying time to market access are that reliable effectiveness data are needed (by prescribers, payers and HTA organizations), demand for stricter safety assessment (by the scientific community and the media) after a series of market withdrawals, and prevention of excess medicalization. Early market access has pitfalls (such as the problem of the toxics) and conditions for granting decisions are not always met (e.g. required studies after granting that are not performed). Granting decisions are always taken under conditions of uncertainty, especially about safety. Higher levels of uncertainty around the risk-benefit ratio are acceptable according to the severity of the disease. There are several regulatory tools for decreasing the uncertainty of safety. Conditional approval can be given for severely debilitating diseases, with additional data requested and penalties for non-compliance. There is the possibility of giving a staggered approval in which prescription is limited to referent centres and/or subgroups of well defined patients. Also a risk management system can be introduced, as was the case with natalizumab. Another possibility is to enhance the tools for safety signal detection from spontaneous reporting systems. The advantage of spontaneous reporting systems is that these can be used by all stakeholders throughout the life cycle of a drug to identify rare adverse events. For that we can use larger databases by data-pooling, add databases in clinical referent centres (not only for efficiency but also for safety), and use patients clinical databases (such as the GPRD in the UK) and patients registries (such as exist in Denmark and Sweden). Pharmacoepidemiology is important in the post-marketing period to improve drug safety. In fact the Agency for Healthcare Research and Quality (Guide for Effectiveness Review, 2007) clearly stated that RCT's are not the best way of analysing drug safety. Only by observational studies the harms can be well estimated. Nowadays there are better methods for the assessment of bias risk on harms reporting as well as common standards for data-mining algorithms. Finally, there is discussion to modify the traditional model of drug development in a way that pharmacovigilance is enhanced.

With regard to new treatments, neurologists are also expecting to be able to provide tailored treatments for each individual patient (personalized medicine). The first way to personalize medicine is to try to select good responders and non-responders. Definitions for response are either clinical (related to relapses, relapse rate and/or disability), MRI-related (T2 lesion load, Gd enhancing lesions) or a combination of these two. One of the best criteria for disability progression as a parameter for response or non-response is the increase of 1 point in EDSS confirmed after 6 months (Rio, 2009). Depending on the chosen criteria for response, the percentage of responders and non-responders can vary considerably.

Personalized treatment requires more information coming from pharmacogenomics. Pharmacogenetics studies genetic variations that result in differences in response to drugs. Genetic information is static for an individual. Pharmacogenomics is the application of genomic technologies to study how a person responds to drugs (genome and its products, RNA and proteins). Pharmacogenomics includes dynamic information. A number of pharmacogenomics studies have been done related to response to INFb. However, we are only at the beginning of studying pharmacogenomics. There are many methodological problems, one of which is the problem of replication of results. Pharmacogenetics and pharmacogenomics can also help us in safety assessment. There is a table of valid genomic biomarkers in the context of approved drug labels. An example in this table is the recommendation for testing for HLA-B*5701 in Asian people at risk for Stevens-Johnson syndrome before prescribing carbamazepine. It is important to include pharmacogenomics studies in RCT's and observational studies of new drugs.

Combination of immunomodulation and neuroprotection

With regard to new treatments, neurologists are also expecting to be able to combine immunomodulation and neuroprotection. It is possible to combine several immunotherapies because their targets are different. But we can also combine immunotherapies with neuroprotection and myelin repair. In that respect there is hope in stem cells. In the past months there have been two important meetings by the MS (STEMS) Consensus Group during which the possibility to launch some trials with stem cell was discussed. From these meetings an article has been written and submitted for publication. In this paper it is mentioned that it is now appropriate to consider an exploratory trial as the next stage using mesenchymal stem cells (MSC's) and neuroprecursor cells (NPC's) in MS. Exploratory studies in SPMS refractory to conventional therapy are recommended with NPC's and MSC's (for proof of principle). This is opening a new way of treating patients differently.

Nowadays MS patients are looking for better treatments, which are easy to use, with better efficacy and at low risk. But they are also asking for reliable and valid information about MS, a more active role in coping with the disease, better communication with their healthcare providers and friends, and easy access to health care services. Patient involvement in therapeutic decision making is very important for discussing the risks of various treatments. Patients vary in how active a role they want to take in decision making. There are autonomous patients who want to make their own decision on the treatment they will receive, but there are also patients who prefer to make an informed choice about their treatment after seriously considering their doctor's opinion. A third group of patients prefers shared decision making in which patient and doctor share responsibility for deciding which treatment is best for the patient. The majority of patients want to be involved in the choice of treatment (informed choice or share decision). 80% of MS patients demand an autonomous role. But it is surprising that their knowledge of risks in terms of absolute numbers is very low. And so it is important to provide patients with clear and visualized information (such as charts) (Heesen, 2007). It has been found that better risk knowledge correlates with a preference for higher autonomy of patients (Kopke, 2009).

But there are other factors that influence patients' decisions, such as patient's risk attitudes and risk perception of the disease (Prosser, 2002) and patients' trust in their physicians (Kraetschmer, 2005). Risk-seeking patients are more likely to choose risky treatments compared to risk-aversive patients. In addition to shared decision making, therapeutic patient education is becoming more and more important.

Finally, some words about the importance of translational research and translational medicine in MS (‘bench to bedside'). Three translational steps can be distinguished. The first translational step is from basic biomedical science to clinical efficacy knowledge. This step is to test whether the treatment works. Key activity is clinical efficacy research. This step works quite well in MS research.

The second translational step is from clinical efficacy knowledge to the clinical effectiveness knowledge. This step is to test who benefits from a promising treatment. Key activities are outcomes research, comparative effectiveness research and health services research. In MS research we need more comparative effectiveness research and observational studies.

The third translational step is from clinical effectiveness knowledge to improved health care quality and value and population health. This step is to test how to deliver high-quality care reliably and in all settings. Key (research) activities are measurement and accountability of health care quality and cost, implementation of interventions and health care system redesign, scaling and spread of effective interventions. In MS research we also need to make progress in this field.

Pathology, pathophysiology and clinical application of new concepts

LASSMANN, Vienna, Austria: Pathophysiology of inflammation and tissue injury in MS: what are the targets for therapy?

There are a lot of data available from RCT's that allow us to judge our concepts of the pathophysiology of MS. Currently, there are three (controversial) concepts of MS pathogenesis. The first concept is that MS is driven by inflammation through all stages of the disease. A second concept (which is often used and is based on clinical and MRI observations) is that MS starts with inflammation (in RRMS) followed by neurodegeneration (in SPMS and PPMS). The third concept (and a rather extreme view) is that MS is a neurodegenerative disease, amplified by inflammation. Pathology learns us that there is profound inflammation, particularly pronounced in the active MS lesions which are most prevalent in the acute and early stages of MS. This inflammation is associated with Gd-enhancement in the active lesions and there is evidence that anti-inflammatory and immunomodulatory treatments are effective in early MS. And importantly, genetic susceptibility for MS is mainly linked to genes that are involved in inflammation.

In the progressive stage of MS (SPMS and PPMS) there is almost no Gd-enhancement, very few white matter lesions and anti-inflammatory therapies are largely ineffective. This is seen as evidence that in this stage there is neurodegeneration which is independent of inflammation. But in the progressive stages of MS all lesions with active demyelination or axonal injury are still associated with inflammation. There are patients where inflammation declines to levels of age-matched controls. So inflammation can die out. This is frequently seen in MS patients older than 60-65 years. In these patients also neurodegeneration (axonal injury) declines to a level that is seen in age-matched controls. So that suggests very strongly that MS related neurodegeneration is driven by inflammation.

The fact that no Gd-enhancement is seen in MRI and no beneficial effect of current immunomodulatory or immunosuppressive treatments can be observed in progressive MS, can be explained by the relation between inflammation and injury of the BBB. In the acute, early stage of MS inflammation is clearly associated with BBB damage. However, in the progressive stage of MS the inflammatory infiltrates of the brain are trapped behind a closed or repaired BBB. This means that absence of Gd-enhancement on MRI does not exclude inflammation in the progressive stage of MS. And obviously, when the BBB is closed, anti-inflammatory drugs are needed that can enter the brain under such conditions to address these inflammatory infiltrates.

In principal there are two stages of MS. The early stage of MS (RRMS) is characterized by new waves of inflammation entering the CNS from the circulation. This is associated with focal demyelinating white matter lesions with variable axonal injury and BBB injury. On the other hand in the progressive stage of MS (SPMS, PPMS) the main substrate is compartmentalized inflammation in the CNS, which is associated with slow expansion of pre-existing white matter lesions, with diffuse inflammation and axonal injury in the NAWM and with extensive cortical demyelination.

And so, from the pathology it can be argued that MS is driven by inflammation in all stages of the disease. The lesional activity is associated with T-cell infiltration. The dominant T-cell population in MS lesions are CD8+ Class I restricted T-cells. Active demyelination and axonal injury is associated with activated macrophages and microglia. T-cells outnumber B-cells in the brain by a factor of 10:1. But still B-cell targeting therapy is quite effective in the disease. The mechanism behind that is still unclear. Is it due to the elimination of antigen presenting B-cells? Or are B-cells a special source of pro-inflammatory cytokines? Or is it perhaps due to the elimination of (EBV) infected B-cells?

There are many mechanisms of neurodegeneration. The most important mechanism that is seen in a vast majority of patients is the neurodegeneration by activated macrophages and microglia cells. It seems to be that there are specific pathways which are of particular importance. This is the pathway of nitric oxide radicals, oxidative damage leading to mitochondrial injury and finally ionic imbalance.

In conclusion, we still think that inflammation remains the prime target for all MS therapies. But clearly, we have to come up with therapies that can reach an inflammatory infiltrate which is trapped within the CNS in the progressive stage of MS. We should consider therapies targeting CD8 T-cells and B-cells, rather than CD4 T-cells. For neuroprotective therapy apparently oxidative damage and mitochondrial injury may become a major target. Stimulation of remyelination may only be effective when inflammation has subsided.

COMI, Milan, Italy: New strategies in pathophysiology: on T-cell traffic and BBB

There are three levels to aim our pharmacological interventions in order to try to get control over the disease: the periphery, the penetration of the BBB and in the CNS. When taking a simplified look at pathophysiology, two mechanisms can be distinguished. The first mechanism is the acute and chronic lesion related damage, the second mechanism is the damage related to diffuse white and grey matter involvement.

In CIS and RRMS the mechanism for neurological dysfunction related to lesion damage is acute, primary axonal damage in new lesions (demyelination and axonal loss). 50% of patients show a residual irreversible increase of disability after a relapse (Hirst, 2008). So when allowing new lesions to appear, we are allowing the accumulation of irreversible axonal damage. From MRI studies we have learned that the amount of lesions accumulated over a period of time, particularly in the early phase of MS, predicts the amount of brain parenchyma fraction (BPF) decrease 13 years later. So something that happened in the beginning of the disease will determine what will happen in the near future.

There is intra-patient variation in lesion pathology. This means that lesions can have a good or a bad recovery at the same time in the same patient. This implies that there is something that has to be locally driven in order to produce this variability. Determining factors are the type and duration of the enhancement and the lesion size. Ring shaped enhancement, longer duration and lesion size larger than 6 mm are correlated with higher risk of a bad evolution of the lesion (‘black holes'). Also lesions located in the periventricular area have a higher risk of evolving in a worse way. One of the explanations is that stem cells, that play a role in the recovery of brain tissue, are confined to the periventricular area. Lesions that occur in this area may severely damage these potentially beneficiary cells.

The reason why large scale, acute axonal loss can remain clinically silent for such a long time is, that the brain is characterized by a redundancy of its organisation. In other words, a lot of axons may get lost without this having any consequences for the functioning of the brain. This can continue until a critical threshold in a given pathway is reached and the compensatory CNS resources are exhausted.

In SPMS (and PPMS?) the mechanism for neurological dysfunction is secondary axonal degeneration in apparently inactive lesions in combination with cortical lesions and diffuse white and gray matter involvement. The evolution of lesion damage is dependent of the phase of the disease. The amount of tissue destruction inside the lesions tends to be minor during the relapsing-remitting stage of the disease. But when the patient enters the progressive stage of the disease, the lesions mature and continue to loose axons in a much more rapid and aggressive manner.

It is clear that the lesions continue to accumulate damage in the progressive phase of the disease. This is partly driven by widespread, diffuse damage of the brain independent of the lesions. There is quite some evidence to substantiate the existence of a degenerative process that is at least partially independent of the accumulation of lesions. Nonetheless, this type of damage seems to be driven, at least partially, by previous lesions.

We have many different drugs that target different levels of the disease (especially periphery and BBB). But we still have to invest a lot of energy to increase our possibilities to target the CNS. This is probably the area where we have to apply our therapeutic efforts to address the progressive phase of the disease.

In conclusion, acute inflammatory lesions have both an acute and a delayed effect on the nervous damage in MS. In these cases early and ‘intensive' anti-inflammatory treatment is highly motivated. Chronic, microglia driven inflammatory activity at the borders of the plaque, expression of a compartmentalized process play some role in the progressive phase of MS. Drugs penetrating the BBB should be tested. There are converging and convincing evidences of a diffuse white and grey matter damage (partially?) independent from lesions. The adoption of preventing neuroprotective strategies should be considered.

EDAN, Rennes, France: Clinical application of new concepts

Axonal damage is driven by inflammation. This concept is supported by the fact that accumulation of disability is well correlated with axonal injury. Also there is a highly significant correlation between inflammation and the extent of axonal injury. The relation between inflammation and neurodegeneration in MS brains is well explained in an article by Frischer, et al (Brain, 2009). One of the key results in their study was that in the progressive stage of the disease active demyelination and neurodegeneration were only seen in patients with pronounced inflammation in the brain. In aged patients at the late stage of the disease the inflammatory process may die out and inflammation declines to levels seen in age-matched controls. If there is a neurodegenerative component which progresses independently from inflammation, one would predict that in such patients axonal injury continues. This was not the case in the mentioned study. In contrast, axonal injury in such patients was similar in extent, compared with age-matched controls.

But this does not resolve the ambiguous relationship between neurodegeneration due to focal inflammation and neurodegeneration due to diffuse inflammation and their relative contribution to clinical deficits occurring during different phases of the disease. A hypothesis with major consequences on therapeutic strategy would be that early focal inflammation might be the pivotal event from which all else follows, leading to consider MS as a two-stage disease. Arguments for this concept are provided by MRI, from a longitudinal MRI study. But also from therapeutic and clinical experience, immunological concepts and epidemiological data on the long-term disability course.

From our study, using data from the Rennes MS EDMUS database, we have found that phase I (EDSS 0-3) and phase II (EDSS 3-6) are strongly independent and that only phase I is dependent on factors influencing the focal inflammatory process: phenotype, early relapses, gender, age of onset. We also demonstrated the concept of MS as a two stage disease with a first stage during which disability progression would be mainly dependent on focal inflammation and a second stage during which disability progression would be mainly dependent on a diffuse inflammation/degenerative process, mainly independent of current focal inflammation.

MS as a two stage disease has clinical applications for new therapeutic strategies. At the later stages of the disease, new medications acting on the diffuse inflammation within the CNS are needed. None of the currently available systemic immunomodulatory or immunosuppressant drugs have showed convincing efficacy. Perhaps fingolimod is a new possibility with a trial ongoing in progressive MS. In the early stage of MS induction treatment strategy to stop the focal inflammatory process formation may be the appropriate concept. We could do so in using few monthly courses of mitoxantrone followed by immunomodulatory agents or a few yearly courses of alemtuzumab.

S1P receptor modulation: from target to treatment

(Integrated Satellite Symposium Novartis: Antel – Vermersch – Havrdova)

ANTEL, Montreal, Canada: Targeting sphingosine receptors in the immune system and the CNS

MS causes inflammation and neurodegeneration, characterized by demyelination, axonal loss, gliosis, and failure of effective repair. Current DMT's target inflammation only. There is a need for novel agents that directly target protection and repair of the CNS as well as targeting inflammation.

Sphingosine 1-phosphate (S1P) is a naturally occurring bioactive sphingolipid that plays a key role in inflammation and repair. Both sphingosine and fingolimod act via S1P receptors. There are five of these S1P receptors (S1P1-5) which are expressed on different cells, either on the immune system or in the CNS. Fingolimod has the capacity to bind to four of the five S1P receptors, which play a role in leukocyte recirculation, neurogenesis and neural cell function, and endothelial cell function, vasoregulation and cardiovascular development.

Before explaining the effect of fingolimod on the immune system, it is important to point out that the effect of this drug is on redistribution of lymphocytes and not on the destruction of lymphocytes. There is a huge pool of lymphocytes in the lymph nodes and there is trafficking of lymphocytes between the lymph nodes and the blood. This trafficking to and from the lymph nodes is mediated by chemokines chemokine-receptor interaction. The S1P receptor is important for the exit from the lymph nodes. The fingolimod effect on this receptor is that the lymphocytes that have entered the lymph nodes will not be able to get out when the S1P receptor is downregulated. In other words, the T-cells are then trapped in the lymph nodes. But only certain cell-types are retained from the circulation. These include the naïve T-cells and central memory T-cells (including Th17), as well as the B-cells. The effector memory T-cells are not affected, because they do not circulate to the lymph nodes and remain in the tissues.

From various studies (both in animals and humans) the following properties of oral fingolimod have been found:

- CD3+ T-cells function is preserved when exposed to fingolimod

- Fingolimod reduces Th17 central memory T-cell (CCR7+) count in the circulation of MS patients. These are the cells that can be held responsible for producing Th17, which has been linked to pro-inflammatory responses in the brain. CCR7 is the marker for the T-cells that travel through the lymph nodes

- Fingolimod changes the proportions of lymphocyte subsets in the peripheral blood: relatively more CD8 T-cells than CD4 T-cells.

- With fingolimod the levels of circulating innate immune cells (such as monocytes and NK-cells) are preserved.

- Fingolimod is present in its active form in the CNS following oral administration

- Fingolimod has prophylactic and therapeutic effects in EAE

- Fingolimod has beneficial direct effects on the CNS in EAE

- Fingolimod has direct effects on progenitor cells and mature human oligodendrocytes in the CNS.

In conclusion, oral fingolimod has a unique mechanism of action, modulating S1P receptors in lymphocytes and neural cells. In the periphery it prevents naïve and central memory lymphocytes to egress from the lymph nodes. This results in only selective T-cell subsets being present in circulation and tissues. These subsets appear to retain their functional properties. Oral fingolimod crosses the BBB into the CNS. Here it may act on S1P receptors expressed on neural cells, modulating functions relevant to MS pathology including endogenous repair mechanisms. It is up to the clinical community to show what the net effect of this drug is, that acts both on the immune system, the BBB and the brain.

VERMERSCH, Lille, France: Visualising treatment effects in multiple sclerosis

Sensitive, convenient and cost-effective techniques are needed to diagnose, monitor and measure treatment effects on the CNS in patients with MS. With MRI inflammation and non-specific pathological changes can be measured. With Gd+ imaging areas of acute inflammation can be detected. T2-imaging and T1 hypointense imaging detect only non-specific pathological changes. The conventional imaging techniques not only lack specificity but are also limited by significant inconvenience, cost and time burden. Valid, reliable and sensitive MRI or other imaging outcomes are needed for the assessment of neuroprotection and repair in MS.

Recently Barkhof and colleagues reviewed 18 imaging measures against 5 criteria (pathological specificity, reproducibility, sensitivity to change, clinical relevance and response to treatment). Imaging techniques included brain MRI (lesional, regional and whole brain), brain PET and functional MRI, spinal cord MRI and visual pathways (including Optical Coherence Tomography, OCT). One of the conclusions was that the assessment of the visual system (anterior visual pathway as indicator for MS) is one of three most promising outcomes to assess neuroprotection and repair. It is clear that the sensitivity to change and treatment response are difficult to assess by analyzing the visual system. For that more long-term follow-up trials are needed.

And so the measurement of neuroprotection by imaging is feasible in MS. It is thought that imaging outcomes in MS provide the best current in vivo measures of neuroprotection and possibly also repair. Among them OCT findings to evaluate the anterior visual pathway. The value of these techniques may lie in diagnosis and monitoring of treatment effects in MS, particularly effects on the CNS and disease progression. But first we need to have proof of concept studies to know whether these promising technologies are also able to detect minor changes over time.

OCT is probably one of the most promising technologies we have today. OCT is a non-contact, non-invasive technique for in vivo examination of the retina. It uses laser light reflection of the retina to create cross-sectional images from a series of laterally adjacent depth-scans. Interestingly, the different cell-types in the retinal layers have different reflectivity and so these layers show significant differences. The result is that you get a histological view of the retina.

With OCT it is possible to measure the retinal nerve fibre layer (RNFL) thickness in the different quadrants of the retina and macula. The RNFL thickness is linked to relapse and disease progression in MS patients. Patients who experienced relapses had a significantly thinner average RNFL compared to relapse-free MS patients over a period of 2 years. Patients who showed disease progression had a significantly thinner temporal RNFL compared to progression-free MS patients over a period of 2 years. Furthermore, RNFL thickness (especially in the lower temporal retinal region) is linked to disease activity and disability in MS patients. It appears that the lower temporal region of the retina is most susceptible of axonal loss. RNFL thickness is also strongly correlated with most of the MRI assessments (especially with the normalized brain volume). RNFL is also a valuable tool for studying optic nerve disease, such as optic neuritis, and probably also in macular pathology, such as in patients with glaucoma or diabetes.

OCT can also measure the macular thickness and macular volume. Eyes of MS patients have a thinner macula than the eyes of healthy controls. Macular thickness may be informative of neurodegeneration in the eyes of MS patients. OCT has shown that macular volume is reduced by 11% average in CIS patients and a history of a single episode of optic neuritis. Further studies are needed to correlate OCT measurements of the macula with MS disease progression.

Conclusion

Emerging treatments directly targeting the CNS demand valid, reliable and sensitive imaging techniques to measure treatment effects on neuroprotection and repair. There are a number of promising imaging outcomes for measuring neuroprotection and possibly also repair, including OCT. OCT is a simple office-based measure that can quantify changes in the retina and may be correlated with MRI measurements of brain atrophy in MS. The latest OCT equipment produces high reproducibility and improved accuracy. It is cheaper, quicker and more convenient for routine monitoring versus MRI, but also complementary to important MRI assessments. OCT introduces a new tool to better characterize the disease process and could prove beneficial in the assessment of new therapies. OCT has already been used as a surrogate marker in phase II trials. Currently it is being used in ongoing phase III and IV DMT studies (fingolimod, GA) in RRMS and PPMS (fingolimod, phase III) and CIS/RRMS (GA, phase IV). There is an OCT advisory board to make sure that this technology is validated and becomes a useful marker for testing and monitoring MS in the near future.

HAVRDOVA, Prague, Czech Republic: Oral fingolimod: the clinical data

Clinical data presented come from one phase II study and three phase III studies. The phase II trial has data after 5 years follow-up. Two phase III studies (TRANSFORMS and FREEDOM) have been completed, the third phase III trial (INFORMS in PPMS patients) is still ongoing. From the phase II and III trials data have become available from quite a number of patients exposed to fingolimod. In the phase II trial 70 patients have had more than 5 years exposure to fingolimod. In the phase III trials 953 patients have received fingolimod for longer than 2 years.

In the phase II study the accumulative number of Gd+ lesions was decreased 43% with fingolimod after 6 months at a dosage of 1.25mg. No differences in number of Gd+ lesions were observed between this lower dosage and the higher dosage (5.0mg). Also it was shown that the time to the first relapse was delayed after 6 months. After 5 years 140 patients (50% of randomized population) remain on fingolimod therapy as part of the phase II trial extension. The effect on inflammatory disease activity appeared to be maintained in patients completing 5 years of oral fingolimod treatment. Over 92% of patients were free from Gd+ lesions after 5 years. Annualized relapse rate was 0.17-0.19 over 5 years treatment, and 61-68% of patients remained relapse-free after 5 years. The most frequent reasons for discontinuation were AE's (16%), withdrawal of consent (15%) and unsatisfactory therapeutic effect (6%). Long-term evaluation has not revealed any safety concerns above those already seen in shorter duration controlled studies.

The efficacy and safety of fingolimod have been and are still being further evaluated in several phase III studies.

The 1 year core study of TRANSFORMS has been completed. In this study 2 dosages of fingolimod (0.5mg and 1.25mg) were compared against INFb-1a in 1,292 RRMS patients. A significant effect on primary and secondary endpoints related to RRMS and significant superiority over INFb-1a was established. Annualized relapse rates of both oral fingolimod dosages were significantly lower than that of INFb-1a (38% and 52%)

The current safety data show that there is a probable link to oral fingolimod pharmacology and reduced heart rate (transient after first dosage and rapidly reversed despite ongoing treatment), macular oedema (in ≤ 1% of patients), generally elevated liver enzyme elevations and blood pressure elevation. There are also some potential risks that require additional clinical trial data or registry data. These include risk of severe infections and risk of skin malignancy.

Also the 2 years core study of FREEDOMS (1.272 patients, 22 countries) has been completed and has shown significant effect on primary (annualized relapse rate) and secondary endpoints (reduction in disability progression based on increase from baseline EDSS, confirmed after 3 months). Fingolimod reduced the annualized relapse rate by 54% for the 0.5mg dosage and 60% for the 1.25 dosage compared with placebo (both p<0.001). In addition, fingolimod slowed the progression of disability by 30% for patients on 0.5mg (p=0.024) and 32% for those on 1.25mg (p=0.017) compared with placebo over 2 years. These findings were supported by positive effects on brain lesions as measured by MRI. The clinical and MRI data will be fully disclosed during the next AAN congress.

The INFORMS study is the only ongoing phase III study in PPMS (fingolimod 1.25mg vs. placebo) at this moment. Patient enrolment is going quite well with nearly half of the required number of patients (650) already randomized. This trial is quite challenging, because you have to find a population that is likely to progress during the trial period. This is why the trial is blinded for 3 years. Measurement of treatment effects is also difficult. This will not only include EDSS, but also the timed 25-feet walk test and the 9-hole peg test. A range of additional techniques is being used to measure treatment effects, such as OCT, brain volume change, MTR, T1 black holes and patient-related outcomes.

Options, challenges and risks of new drugs

KAPPOS, Basel, Switzerland: What to do in the Clinical Isolated Syndrome?

There are several advantages of early intervention in CIS, which are not all supported by evidence. Epitope spreading and compartmentalization into the CNS would make the auto-immune process more difficult to handle, especially with currently available agents. Then there is the decreasing role of (treatable) immune pathology as the disease goes on. By avoiding permanent damage and diminishing sub-lethal hits the potential of regeneration and reorganization would be better preserved.

Evidence of early damage in CNS tissue in patients with CIS comes from neuropathology (early axonal loss associated to inflammation (Kuhlmann, 2002)), MRI (early axonal loss, changes indicating early matrix destruction (Filippi, 2003)) and neuropsychology (early subtle deficits). Serial MRI studies show ongoing subclinical inflammatory activity.

Evidence supports the prognostic relevance for long-term disease progression of those disease measures that are able to modify the natural course of disease with immunomodulatory treatments, in terms of time from first to second relapse, initial relapse rate (in the first 2 or 5 years), disease progression in the first 5 years and baseline T2-lesion volume. Changing these parameters does not automatically result in a change of prognosis.

The main challenge for CIS treatment is to assess the individual MS-related risk. The prognostic value of MRI in early MS is undisputed on a population level, but unfortunately not sufficiently enough on an individual level. Long-term follow-up studies have shown that markers for bad prognosis do not mean everything on an individual level. A study has shown that almost 20% of CIS patients with 10 or more T2 lesions at baseline remained CIS after 20 years (Fisniku, 2008). New MRI methods (such as T1 black holes, atrophy, spectroscopy and fMRI) look promising in adding to the prognostic value of MRI, but they need further validation.

In neurophysiology evoked potentials are a valid measure of functional damage, but this measure lacks long-term follow-up for prognostic evaluation. Analysis of body fluids (CSF, blood) does not provide risk factors that predict disease prognosis at an individual level.

Next challenge is to determine the impact of early treatment on the conversion to clinically definite MS (CDMS) and on the progression of sustained impairment and disability. There have been four controlled studies in CIS with INFb and GA (ETOMS, CHAMPS, BENEFIT and PRECISE). The main finding from all these studies is that it is possible to delay CDMS, even though methodological differences make it difficult to assess outcomes across these studies. From these studies it is not possible to derive a clear picture of the factors that determine the efficacy of treatment in CIS. Younger age, number of Gd+ lesions and T2 lesions at baseline are indicative for response rate. Information about the effect on delay of disease progression stems mostly from the BENEFIT study, that was prospectively designed to assess the effect on early MS. In the 5 years analysis the enthusiasm about the long-term effect of early treatment vs. delayed treatment was dampened. After 5 years the clinical differences found in favour of early treatment still existed, but these were no longer statistically significant. One of the reasons was that the EDSS score did not change much in the total group over the five years period and remained at a low level. The only outcome after 5 years that pointed in favour of early treatment was cognition (PASAT). Improvement in PASAT scores was more robust and pronounced in the early treatment group. Whether this is just an isolated finding or a real finding still remains an issue open for debate. The conclusions from the BENEFIT trial are that the relative benefits of initiating treatment immediately after the first episode of symptoms highly suggestive for MS versus a delay by several months up to 2 years tend to decrease with the length of observation and ongoing treatment. Over an observational period of 5 years, progression of disability was very mild in both treatment groups. Persistence of statistically significant favourable effects on the rate of conversion to clinically definite MS, on inflammatory disease activity in MRI and on cognitive performance after 5 years, support early initiation of immunomodulatory treatment.

Final challenge is to determine the substance to use and the preferred strategy. Should we start early with the highest tolerated dose and the strongest compound or start low and then escalate? And then it is important to determine which of the new compounds in phase III will be candidates for early treatment. We know that with some of these substances CIS studies have already started. This will also raise the dilemma whether to choose for drugs that have the lowest expectation of long-term important serious adverse events or to choose for drugs that promise an immediate and a more intense affinity. These are questions that need to be further explored.

In conclusion, we should offer the option of early treatment to all patients with CIS (who fulfil the pivotal trial criteria), and insist in those with higher risk. We should explore the impact on disability with better measures and long-term follow-up, and explore impact of more aggressive treatments and strategies.

VERMERSCH, Lille, France: What to do in fulminant multiple sclerosis?

There are neither guidelines nor consensus about what to do in fulminant MS. Also there is no clear definition of fulminant MS. Fulminant MS belongs to the spectrum of demyelinating disorders with very acute onset and usually a severe clinical course. There are two conditions of fulminant MS. One is a rapid sequence of exacerbations, with intervals between bouts less than 30 days. Most of the cases are single ‘catastrophic' clinical manifestations with MRI appearance of large tumour-like (tumefactive) lesions. This is the typical fulminant monophasic Marburg-type of MS.

Marburg-type MS is part of fulminant idiopathic inflammatory demyelinating diseases and is characterized by the absence of previous neurological symptoms, a fulminant course and at necropsy evidence of extensive axonal loss, and necrosis. Usually these patients have an impaired vision or may be comatose and have global encephalopathy (with seizures, aphasia, severe motor impairment). It may be fatal due to an extension of the lesion in the brainstem. On MRI these patients may show very large T2 lesions and (ring shaped) Gd+ lesions. Sometimes multifocal T2 lesions are observed or massive lesions in the cerebellum or brain stem. And sometimes also older T2 lesions are observed, suggesting a history of MS pathology.

It is important to differentiate fulminant MS from other severe conditions, such as highly active RRMS, ADEM, Balo, neuromyelitis optica and systemic diseases with neurological involvement (such as Behcet's disease).

Treatment methods for fulminant MS include steroids, plasma exchanges, cytotoxic drugs, rituximab, haematological stem cell transplantation (HSCT) and decompressive hemicraniectomy (in rare cases).

There is a clear relationship between humoral pathological changes and response to therapeutic plasma exchange. In all MS cases with pathological pattern II plasma exchange proved to be successful (Keegan, 2005). Treatment failure was only seen in MS cases with pathological pattern I or III. Early initiation of plasma exchange and improvement at discharge are predictors of positive response to treatment at 6 months. The effects of rituximab, HSCT and decompressive hemicraniectomy have been supported by positive case findings.

Brain biopsy is recommended in most cases of fulminant MS. It allows pathology-directed immunotherapy, but is also of importance in differential diagnosis with glioma, multifocal glioma, gliomatosis cerebri and lymphoma (sarcoidosis). In a minority of cases biopsy in not recommended. For instance if there is a clinical history of a previous and highly suggestive event or if there are multifocal lesions of different ages with a positive CSF. Brain biopsy is needed in case of negative outcomes after plasma exchange.

The proposed treatment strategy in fulminant MS is as follows. Upon presentation of the patient in the neurology unit (or intensive care unit) and after diagnosis of fulminant MS one is to start with a megadose of steroids (IVMP 1-5g/day during 5-10 days, then tapering oral steroids). In case of a positive outcome an MRI is then made after 30 days. If this MRI shows no Gd+ lesions and a decrease of T2 lesion load, the patient then is followed-up clinically and with MRI (e.g. after 2 months). Depending on the condition of the patient a DMT treatment (with IFNb or GA) should be considered. If MRI after 30 days does show Gd+ lesions or an increase of T2 lesion load, mitoxantrone should be given for 6 months as an induction therapy followed by GA or INFb.

In case of a negative outcome after 15 days of steroids treatment, plasma exchange should be given, followed by 6 months of mitoxantrone or rituximab (in case of pattern II MS pathology after biopsy). In case of a positive outcome of this treatment, a follow-up is given with GA or INFb. In case of a negative outcome after mitoxantrone or rituximab, HSCT should be discussed.

HARTUNG, Düsseldorf, Germany: The role of new drugs in the treatment of relapsing remitting multiple sclerosis

The new drugs for the treatment of RRMS that are now in phase II and III of clinical development can be classified by their targets. These targets are migration (fingolimod), T-cells (alemtuzumab), B-cells (rituximab, ocrelizumab, ofatumumab, atacicept), immunosuppression (laquinimod, cladribine, teriflunomide) and neuroprotection (BG-12). We have made great advances in our understanding of the immunopathogenic cascade. The ultimate goal is to use that knowledge to design effective and safe drugs.

Fingolimod is a mimic of a metabolite that binds to S1P receptors distributed throughout the human body. Fingolimod may combine immunomodulatory and reparative properties. It may address two key features of the pathological process of MS. First by modulating the immune response in the periphery and maybe also in the CNS by inhibiting access of pathogenically active T-cells, but also in modulating neural cells that are involved in the pathogenesis of MS. For mode of action and clinical development results see lectures by Antel and Havrdova.

Therapeutic monoclonal antibodies have been devised with the view that they may provide an exquisite specificity and targeting either towards cells of essential importance to the pathogenic process or molecules that are at checkpoints of the immunopathogenic cascade. Monoclonal antibodies have a number of advantages for therapeutic use. They offer exquisite specificity, appear to generate increased efficacy and they are infrequently administered. However, there are also caveats that need to be taken into account and continuously reassessed.

These caveats are the functional consequences off-target of sustained T-cell and B-cell depletion or silencing. Can protective immune responses be mounted?Also unexpected, rare but serious side effects may occur (PML, malignancies), which may be due to impaired immune surveillance. We clearly have to keep this in mind when we are considering a change in our therapeutic algorithm in RRMS.

Alemtuzumab is a humanized monoclonal antibody directed at a glycosylated, GPI anchored protein that has a cytotoxic effect and reduces circulating T-cells and B-cells, but also monocytes, NK cells and dendritic cells. It has a very prolonged effect by reconstituting a new lymphocyte pool. There is a significant and sustained impact on circulating T-cells, whereas B-cells tend to recover quite early. Results from the phase II CAMMS trial were quite impressive, showing a profound risk reduction in sustained accumulation of disability (i.e. 1 point EDSS increase sustained for 6 months) of 71% and a reduction of annualized relapse rate of 74% with alemtuzumab vs. INFb-1a. These effects were sustained for a longer period of time (48 months). Serious adverse events were registered with alemtuzumab related to the emergence of auto-immune diseases, such as ITP and auto-immune thyroid disorders.

Rituximab is a humanized antibody directed against CD20, which induces B-cell depletion without affecting plasma cells. There have been two RCT's exploring therapeutic utility of rituximab both in RRMS and PPMS (with negative outcome). In the phase II RRMS trial the primary outcomes related to MRI activity were met with rituximab, producing a significant decrease in MRI activity, both early on and sustained.

Cladribine is a synthetic purine nucleoside analogue with the capacity to cross the BBB. It apparently causes preferential and sustained reduction of lymphocyte numbers and has a greater effect on CD4+ than on CD8+ T-cells. Other haematological and immune cells, including B-cells, are relatively spared. The exact mode of action still needs to be explored. There is evidence that it reduces levels of pro-inflammatory cytokines and chemokines and impedes T-cell and monocyte migration. Cladribine causes a sustained reduction in T-cell counts. The phase III CLARITY trial has been communicated but not yet published. Cladribine was shown to significantly decrease the annualized relapse rate (55% and 58%) compared to placebo. It also decreased the disability progression (EDSS) by 33% (low dose) and 31% (high dose) and consistently impacted on MRI activity (Gd+ lesions and T2 lesions). Also the disease activity-free status (combining various aspects of disease activity) after 96 weeks was clearly superior with use of cladribine vs. placebo. Cladribine was generally well tolerated. The most commonly observed AE was lymphopenia (26.7%). Cases of cutaneous herpes zoster were reported (2.3%), but no opportunistic, serious infections. In the cladribine group were 3 cases of solid tumours (melanoma, ovarian carcinoma, pancreatic carcinoma) and 1 case of precancerous cervical dysplasia.

Fumaric acid (BG-12) has a large body of experience in dermatology and is well tolerated. It has a potentially new mode of action, targeting a pathway (Nrf2-pathway) that may be critically involved in axonal damage, though this needs to be further substantiated. In the phase III trial 3 different doses of the drug were compared with placebo. The results showed a clear reduction of new Gd+ lesions, new T2 lesions and new T1 lesions with the drug. A large phase III trial is ongoing.

In conclusion, new therapies will become available very soon. In all likelihood we will see two players, cladribine and fingolimod, become approved by 2010. Others are currently in phase III (laquinimod, teriflunomide, BG-12). Alemtuzumab and rituximab may have a favourable dosing regimen. But clearly it would be premature to expect an immediate abandonment of our long established first line drugs. Not that it wouldn't be justified, based on the pivotal trials and looking at the clinical outcomes, but because of safety concerns as a reflection of the natalizumab experience. With more powerful immuno-interventions rigorous vigilance and risk management programs are mandatory for the early detection and handling of serious events. Only through experience we will be able to readdress the therapeutic algorithm.

MONTALBAN, Barcelona, Spain: The role of new treatments in the progressive forms of multiple sclerosis: SPMS and PPMS

There are no treatments available for PPMS and only a few for SPMS (IFNb-1a, IFNb-1b and mitoxantrone) in patients with superimposed relapses. For SPMS patients without superimposed relapses there is no treatment.

There are three reasons why drugs fail in progressive MS. The first is that pathogenic mechanisms in the progressive phase are completely different from those in the relapsing phase of MS. Second reason is that patients populations included in the trials are not appropriate and the third is that clinical outcomes are not sensitive enough to detect worsening of the disease over a given period of time.

Looking at pathogenic mechanisms in progressive MS, there are focal inflammatory demyelinating white matter lesions which are similar to those found in patients with RRMS. But in addition there is diffuse damage of the NAWM and demyelinating lesions in the gray matter, in particular in the cerebral and cerebellar cortex. In progressive MS inflammation is trapped behind a closed BBB, and damage of the CNS parenchyma is provoked by the action of diffusible factors acting on microglia cells and a few intraparenchymal T-cells. The failure of current immunosuppressive or immunomodulatory treatments in patients with SPMS and PPMS may, thus, be more related to their inability to pass the BBB and to reach therapeutically relevant concentrations within the CNS compartment.

Some positive signs for future treatment possibilities of progressive MS are coming from oral fingolimod (see previous lectures by Antel and Havrdova) and fumaric acid (BG-12). BG-12 may ‘protect' oligodendrocytes and neurons from inflammatory and metabolic damage via activation of Nrf2 and of the antioxidant and metabolic defense mechanisms. And perhaps other possibilities will emerge from the new MS treatments that are becoming available in coming years.

From a series of three studies recommendations have been drawn to provide more sensitive clinical endpoints in PPMS trials. It has been suggested to use a combination of worsening of EDSS and timed 25-feet walk test. The time of observation should at least be 2 years, and a cut-off of 20% for the 9-hole peg test seems desirable.

Recommendations for patient selection for PPMS clinical trials are to include younger patients with shorter disease duration (less than 10 years), with lower EDSS values and with active disease (measured by changes in EDSS in the previous year) and active scans.

Laquinimod: A novel oral immunomodulator in development for the treatment of multiple sclerosis (Integrated Satellite Symposium Teva: Brück – Comi – Kieseier)

BRŰCK, Göttingen, Germany: Insights into laquinimod mechanism of action

Laquinimod is a novel, once-daily, oral, immunomodulating agent currently under development for treatment of RRMS in two phase III clinical trials. It is able to cross an intact BBB and is detectable in the brain of both EAE and normal mice. Laquinimod does not affect cell viability or cell proliferation in high doses and for long-term exposure. So it is not acting by immunosuppression, but by immunomodulation.

The following effects of laquinimod on inflammation have been observed in animal models:

- Laquinimod has a dose-dependent effect in prevention of EAE in MOG-induced mice

- Laquinimod has a therapeutic effect on EAE in MOG-induced mice with a clear clinical improvement of the disease

- In the preventative approach of EAE laquinimod significantly inhibits inflammatory cells (both T-cells and macrophages) from entering the CNS

- In the therapeutic approach of EAE laquinimod significantly inhibits macrophage infiltration into the CNS, without significant inhibition of T-cell entry into the lesions or the anterior white matter.

- Laquinimod induces a shift from Th1 to Th2 cytokines in Lewis rat EAE models, both in the periphery and in the CNS. Laquinimod inhibited expression of IL-12 and TNFa mRNA, and upregulated expression of Th2 and Th3 cytokines IL-4, IL-10 and TGFb mRNA in spleen cells. In spinal cord sections laquinimod inhibited expression of TNFa mRNA, and upregulated expression of TGFb mRNA.

- Laquinimod reduces the production of Th1 and Th17 pro-inflammatory cytokines (IFNg, TNFa, IL-6, IL-17 and IL-13) in MOG EAE

- Laquinimod efficacy is IFNb independent. Laquinimod dose-dependently inhibited disease development in chronic EAE in both IFNb knock-out mice and in wild type mice.

In animal models preventive and therapeutic treatment with laquinimod was not only proven to be effective in ameliorating the disease severity and extent of inflammation, but also in reducing demyelination and axonal damage. In MOG-induced EAE mice preventive and therapeutic treatment with laquinimod was shown to reduce spinal cord demyelination and axonal loss.

In summary, laquinimod is an immunomodulator with both anti-inflammatory and neuroprotective properties. Laquinimod induces a shift in cytokine balance towards a Th2/Th3 profile, it reduces immune cell infiltration into the CNS and induces myelin and axonal preservation. Laquinimod is able to reach the CNS, so it may not only act in the periphery but also within the lesions in the CNS. Laquinimod's mode of action supports the emerging clinical data pointing to an excellent risk-benefit ratio. The mode of action is further explored in the phase III clinical programme. Within these studies a lot of parallel experimental studies will go on concerning PBMC analysis and gene expression, monocyte analysis, dendritic cells analysis, cytokine measurement in CSF and blood, T-cell responses in CSF and biomarkers for neurodegeneration and inflammation.

COMI, Milan, Italy: Efficacy and safety of laquinimod: current data and future perspectives

Laquinimod is a novel oral, once-daily immunomodulator, currently in phase III development for RRMS. Most of the clinical efficacy and safety data are derived from 2 Phase II, double-blind, placebo-controlled studies. This already provides us with a clear indication of the potential of this drug. There are 2 major phase III RCT's ongoing (ALLEGRO and BRAVO) and we have to await the results in order to state something definitely. Based on the mechanism of function of this drug no major adverse effects should be expected.

The first phase II study (ABR-01506203) was double-blinded and placebo-controlled. Patients (RRMS or PPMS, 18-65 years, EDSS 0-5.5) entering the study had to have an active disease as revealed by clinical or MRI measures (or a combination). Primary endpoint was the cumulative number of active lesions (Gd+ or new T2 lesions) at week 24. The Gd was administered in a triple dose. Of the 209 patients randomized on laquinimod (0.1mg or 0.3 mg) or placebo, about 95% of patients completed the trial. No difference in treatment effect was observed between the daily dose of 0.1mg laquinomid and placebo on the primary endpoint. However, in the 0.3mg dose there were some evidences that the drug was able to reduce the accumulation of disease activity.

And so the conclusion of this study was that the 0.3mg dose showed encouraging effects on the evolution of inflammatory lesions in this cohort of patients. Also with an excellent safety profile, and assuming a dose-dependent effect, it was decided to conduct an additional Phase II study, but with a higher dose, for a longer duration.

The subsequently performed phase IIb study (LAQ/5062) was also double-blinded and placebo-controlled. Patients (RRMS, 18-50 years, EDSS 1-5) had to have an active disease (≥1 relapse in the year prior to screening or ≥1 Gd+ lesion on MRI screening). Primary endpoint was the cumulative number of Gd+ lesions on MRI scans in weeks 24, 28, 32 and 36 (last 4 scans). This time Gd was administered in a single dose. Of the 306 patients randomized on laquinimod (0.3mg or 0.6 mg) or placebo, about 92% of patients completed the trial. This time no statistically significant effect on any of the predefined outcome measures was observed with the 0.3mg dose. However, a quite consistent, homogeneous and statistically significant effect was observed in the 0.6mg group both on primary and secondary outcome measures. In post-hoc analysis evidence was found of a 30% reduction in the deterioration of brain atrophy on 0.6mg compared to placebo. This is an interesting finding, considering that a drug reducing inflammatory activity also has some single atrophy effect. This may open the possibility that the drug is also doing something at the level of neuroprotection. Despite the short duration of the trial, the annualized relapse rate showed a trend towards reduction in the 0.6mg group. Concerning safety and tolerability no difference were observed in frequency of AE's and SAE's between laquinimod and placebo. In each group only 1 possibly drug related SAE was observed. This concerned a case of menometrorrhagia on laquinimod 0.6mg and a case of elevated liver enzymes on 0.3mg. Observed cases of liver enzyme elevation were all reversible, normalizing in most patients while remaining on laquinimod. In addition there were no signs of liver damage or failure and no concomitant bilirubinemia.

In a double-blinded extension of this phase IIb study all placebo-treated patients were randomized either on laquinimod 0.6mg or 0.3mg. Unfortunately, some patients had a temporary interruption of treatment (with an average of 20 days) before entering the extension study. It turned out that the positive effect of laquinimod on MRI active lesions was reproduced in patients who had switched from placebo to active treatment. The effect on MRI active lesions was sustained throughout the treatment period. One of the reasons for this extension study was to monitor the long-term safety and tolerability of the drug. Findings were that with increased exposure in the extension study, there was a decrease in the incidence rates of elevations in liver enzymes. No opportunistic or life-threatening infections were reported in patients treated with laquinimod. No cardiac events, suggestive of myocardial ischemia were reported in patients treated with laquinimod and no increased rates of specific malignant diseases.

In a third step observation level of the phase IIb study, all patients were given 0.6mg laquinimod in an open label 2 years extension. So this provides us with 3.5 years of follow-up data on laquinimod. From this extension a persistent decrease in MRI activity by 1/3^rd has been observed overall. In addition the percentage of Gd+ and

T1 lesions free patients has increased by 60% compared to baseline after the 2 years open-label extension.

In conclusion, based on the consistent evidence from the phase II studies, laquinimod is a drug with a very good possibility of becoming a major player in the area of MS treatment in the near future. The ongoing phase III studies ALLEGRO and BRAVO will provide us with the definite evidence.

KIESEIER, Düsseldorf, Germany: Making the right treatment decision. Importance of evaluating the risk-benefit ratio of multiple sclerosis treatment in the new era

What are the choices that we have in MS? Are there any kind of guidelines that we have to establish or ways of new thinking that we should implement in our way of treating MS patients? Current and near future MS therapies provide very promising benefits but also growing concerns about safety. And so the risk-benefit ratio is top of everyone's mind in MS therapy. The question is then how to look at risk-benefit ratio in order to make our treatment decisions and what are the factors that need to be considered?

Step one is how to define ‘benefit'. From a patient's perspective benefit is defined in terms of disability, soft symptoms (such as fatigue, depression, cognitive deficits), quality of life, family role, social impact and impact on career and jobs). But that is very hard to put in a solid rock perspective in terms of making judgements. So we look at the scientific perspective of benefit. We have the primary outcomes of the trial (reduction of disease progression, reduction in relapse rate). Freedom of disease activity might be a new way of looking at the efficacy of a drug, as well as the improvement of physical disability (EDSS improvement). So provocatively it can be said that in the old days we were looking into the failure-based paradigm (aimed at trying to prevent disease progression), and with the novel drugs we are moving on into the improvement paradigm by trying to improve disability.

When defining the benefit of a specific treatment it is invalid to compare different trials due to known and unknown differences between responses of the active and placebo group. Additional head-to-head data are needed to be able to compare different therapies. These data may become available in the near future, including some of the emerging MS agents not yet approved. So at present the burden falls upon the clinician and the patient to make a decision together as to the optimal treatment for each individual.

Step two is the risk awareness and assessment. Risk assessment of new treatments is quite difficult. At the time of approval the knowledge of a new drug is incomplete, especially with reference to its safety profile, due to a variety of factors including constraints in sample size. We have to be aware that particularly biologicals carry specific risks. They are derived from living sources and their production and purification are very delicate. The predictability of preclinical data to humans is limited for biologicals, specifically concerning safety, due to the species-specific action and immunogenic properties in animals.

In a recent study the nature, frequency and timing of safety-related regulatory actions were determined for biologicals following approval in the US and/or EU. The 174 biologicals studied had a probability of 14% that a first safety related regulatory action would be required within 3 years after receiving marketing approval. This probability increased to 29% with 10 years after approval. The major warnings issued were based on the immunomodulatory effect of many of these biologicals. This shows that health care professionals must be aware of the specific risks related to new treatments to be able to provide a link between the use of the new drug and the patient presenting with a clinical problem. Therefore close monitoring is recommended.

In assessing the risks of a new treatment, two variables need to be taken into account: the number of patients exposed to the drug and the duration of the overall exposure. The challenge is to evaluate the risk-benefit ratio for an individual patient relying on data from different study populations. In particular in MS, the lack of sufficiently good prognostic markers and the lack of risk factors for SAE's makes it even more difficult to individualize the risk-benefit ratio for the individual patient.

And so there are several reasons for concern. There is a lack of surrogate markers for SAE's and a lack of prognostic factors helping to select the right patients. The effects for long-term immunosuppression are unknown, as are the effects of highly selective immunosuppression. The exposure rates to new approved drugs are limited, and thus the real risks are not predictable. This is particularly relevant in MS, since it is a chronic disease that requires long-term treatment. Because also young women at childbearing age are treated, it is relevant to have insights into the reproductive toxicity of any new drug.

The question now is whether it is already time for a new way of thinking and treating. If risks of a drug are time- and/or dose-dependent, should we implement deescalating therapies? For example mitoxantrone shows dose-dependent cardiotoxicity and natalizumab has a dose- and time-dependent risk for PML.

Making the right decision means, that on the drug side a number of risks and benefits are to be taken into account: efficacy (short and long-term), safety, tolerability, convenience and costs. The best approach is to minimize the risks and maximize the benefits of new drugs. This means making a careful patient selection according to disease course and activity, individual evaluation of other suitable treatments with better risk-benefit ratio for this specific patient, and considerations on the overall treatment strategy for this patient including the potential necessity for future treatment escalations. In general drugs with low risks are to be preferred. For new treatments risk management programs and pharmacovigilance are very important. And patients need to be educated so that they are aware of what they are doing.

In summary, it is extremely difficult to assess the clinical benefits of a new drug in comparison to established treatment strategies as long as head-to-head trials have not been conducted. To assess the real risk of a new drug is even more difficult as long as data on long-term exposure and high numbers of patients being treated are lacking. The burden falls upon the clinician and the patient to make a decision together as to the optimal treatment for each individual.

Response measurement and how to use the tools. MRI, CSF, biomarkers, pharmacogenomics

FREEDMAN, Ottawa, Canada: How to evaluate the treatment response?

At the moment we are only able to really treat focal inflammation in the early part of MS. That period is called the time window of opportunity for early treatment. So knowing about the medications, including the ones that are coming, you have to consider where your patients are in this window when you first see them. The type of approach and the treatment response depends on where in the window the patient is when initiating treatment. When dealing with a new disease presenting early, you can start with medication that is well known and use an escalating approach. But when the patient is already nearing the end of the window of opportunity, the approach is going to be more aggressive because the window is going to close shortly.

Questions that come up when trying to define response to treatment in the absence of a cure for MS, are how to define ‘sub-optimal' response to treatment and how to decide which aspects of residual disease activity are worrisome and which are not.

The response to currently available medications is usually very good. So if patients do not respond to drug treatment this can be because they do not have the right diagnosis (e.g. NMO instead of MS) or that patients are not taking the drug as prescribed (e.g. due to intolerance issues or dose modifications after abnormal lab test results). And then it is also important to determine the nature of this suboptimal response and whether the presented complaints are related to attack. There are many symptoms (like mild sensory attacks, bladder issues, or paroxysmal symptoms) that are often misconstrued as related to attack. And the best way of determining this, is by assessing patients in hospital when they present complaints suggestive of an attack. Probably the number one reason for non-response is that the patient's disease is advancing. And perhaps they are already into the progressive phase of disease. And if the disease is advancing then that is the number one reason why a change in treatment, possibly an escalation, needs to be considered. And then there is a series of biological and other factors (e.g. NAB's) that make some patient better responders to DMD treatment than others.

There are several measures to be used in assessing treatment response. These include relapse, progression and MRI. But the question is whether the same outcomes as those used in clinical trials can also be used in modifying medications and monitoring treatment response. With regard to relapse it is important not just to consider rate, but also other aspects such as severity and recovery. In fact, relapse rate is totally undependable. The only way of knowing what kind of relapse patients are experiencing is by bringing them in and assess it. And then bring them back in after a while to assess the residual effect after the episode of attack.

The considerable drop in annualized relapse rate seen in contemporary DMT studies (e.g. with natalizumab in comparison to INFb in a trial conducted a decade earlier) can be largely explained by patient selection. The earlier studies were conducted in patients with a more advanced state of MS, whereas contemporary studies are conducted with patients in the early stage of disease. As a result all the older drugs (such as INFb and GA) are showing a better response in today's populations.

Some years ago the Canadian MS Working Group (CMSWG) issued recommendations on optimizing treatment response with DMD (Freedman, 2004). This also involved determining which patients are getting a true benefit from the treatment. These recommendations were based on the literature at that time, but hardly need revision based on today's knowledge. The levels of concern (low, medium, high) were defined to consider treatment modification based on relapse, progression and MRI outcome criteria. High levels of concern indicated a clear need for treatment change, whereas medium levels indicated that the patient should be monitored more closely in order to decide whether a treatment change might be necessary.

Annualized relapse rate is totally unsuitable to communicate to patients, because a 30% or 60% reduction doesn't mean anything to them. Patients are mainly interested in looking at the situation before and after the start of treatment. A good deal of the observed reduction in annualized relapse rates during the trials has been due to use of placebo anyway (to as much as 76% in the cladribine trial). So how much of the relapse regression will be caused by the drug effect is very hard to tell in an individual case. This is why relapse rates present a problem to evaluate. Even the relative rate reduction of the drugs does not provide much to go on. From the trials in which patients showed similar baseline relapse rates, we know that this doesn't mean that all these patients run a similar risk of an event. The observed differences in placebo rates between these patients have told us that. The low rates of risk reduction in relapse and progression found in contemporary studies make it necessary to think of a new matrix of validated measurements, because the current way of evaluating (i.e. by relapse rates) is running short. Critics say that relapse counts are inflated (most clinical studies focus only on the number of (severe) attacks that change EDSS scores), many attacks are missed (attack rates depend on the contact rates and attacks with profound fatigue, minor but persistent sensory signs and significant cognitive change are missed). Still, relapse rate matters, because it is correlated to disability, in such a way that each additional relapse increases the odds of progressing by approximately 40%. But also the relapse quality needs to be considered, because this predicts residual effect (e.g. severity of attacks, number and location of lesions in the CNS, particularly in the cerebellum).

When looking at progression it is clear that baseline EDSS confers to risk of progression. EDSS has been proven a useful indicator for sustained disease progression. EDSS changes should reflect changes in the same Kurtzke functional subscore (KFS). If not, this tells you that you may have missed an attack. Changes in multiple KFS are more predictive of progression. First year progression after the start of DMD may be more reflective of the inflammatory events in the previous year, so events in the second year are more indicative of the current treatment effect. KFS changes in bladder or vision are unlikely to impact on overall EDSS progression (these are downgraded in the current NeuroStatus).

With regard to MRI there is a clear correlation between MRI findings and numbers of relapses. However, more than 10 new T2 lesions on MRI over a 2 year period are needed in order to matter in terms of relapse rate or progression. The CMSWG has not defined a level of high concern based on MRI findings. According to this working group there is not a single change on MRI that is absolutely required to indicate a change in therapy. That is because it is hard to define which of the MRI measures are of more importance than others.

The CMSWG designed a model that assesses the concern whether to modify a treatment regimen based on relapse, progression and MRI. In case of 3 low levels, any 2 medium levels or any 1 high level of concern, this is seen as an indication of suboptimal treatment that might warrant a change in therapy.

As the number of clinical events diminish we might rely on a disease free composite, of being free of any activity in terms of relapse, progression or MRI. The difficulty will be to decide on how much weighting to give to each of these outcomes. Are all relapses, progressions and MRI lesions the same? Clearly a single sustained progression is probably worth x relapses and y MRI lesions, but evidence is lacking to decide what values to assign x and y.

In conclusion, the treatment response in the first 12-18 months is highly predictive of future response. Disease activity can be judged using the same outcome measures of relapse, progression and MRI, keeping in mind issues such as quality of the events. Suboptimal responders should be switched to a different therapy (the window is narrowing). We may be ready to aim for truly being ‘disease-free'.

BARKHOF, Amsterdam, The Netherlands: The use of MRI in treatment response studies. Markers for inflammation, neuroprotection and repair

In inflammation MRI has an accepted role and proven efficiency in RCT's with Gd-enhancement and new T2 lesions. Still sensitivity needs to be increased and analysis of data needs to improve when faced with populations with less active disease or in the monitoring of active comparators. In the area of neuroprotection and repair there is an emerging role for MRI in treatment monitoring.

Longitudinal MRI monitoring with Gd-enhancement in MS patients with active disease has learned us that disease activity (e.g. new lesions) continues without patients having to experience relapses. It can occur that a new relapse is preceded by a bout in MRI-activity. So by monitoring MS treatment with MRI we can observe disease activity (such as frequency of new lesions) which may redefine the natural course of MS. The relevance of this subclinical activity is still unclear. With MRI we can also evaluate anti-inflammatory treatment and use it as a primary outcome in phase II studies in which MRI offers a more rapid and efficient screening of new therapies. But also MRI can be used as supportive evidence in phase III trials and to do subgroup analysis. It is much more difficult to interpret MRI findings for individual patient implications. At a trial population level, however, a clear linear trend between the occurrence of relapses and MRI activity has been observed (Sormani, 2009).

The use of Gd has a downside in that it may cause nephrogenic systemic fibrosis (NSF) as a side effect in patients with poor renal function. Gd remaining in the tissues for months can induce fibrosis that can be quite severe. This is one of the reasons why patients are screened on renal function before receiving contrast. But more generally we have to ask ourselves if we really need all this Gd, and be aware of repeated dosing in trials. It is just as well to evaluate the occurrence of new T2 lesions by comparing follow-up scans. One way of improving the sensitivity of this measure is by method of registration and subtraction. After registration of the MRI images their differences are noted by subtraction on the computer. When performing a power calculation to design a study and looking at the patient numbers needed to get an 80% power to discern treatment effect, it is surprising to find out that the number of subjects needed is lower with this T2 subtraction technique than with Gd-enhancement. The number of MRI scans needed is even considerably lower with T2 subtraction than with Gd-enhancement. So this may be a useful technique for the future that may also help us detect smaller treatment effects.

For neuroprotection and repair monitoring some good imaging candidates are available, but these certainly need more proof-of-concept studies. About one year ago a workshop was held to review the imaging outcomes for neuroprotection and repair. It was decided that the most promising measures to be used in phase II trials are changes in whole brain volume to gauge general cerebral atrophy, T1 hypointensity (‘black holes') and MTR to monitor the evolution of lesion damage (remyelination and axonal loss), and OCT findings which are related to atrophy status.

Power calculations show that these outcome measures can be applied with attainable sample sizes. General brain atrophy is a powerful tool to evaluate the neuroprotective effect of treatment. This also applies to OCT and MTR. MTR is quantitative and sensitive and allows detection of even smaller treatment effects over shorter periods of time. Persistent black holes (PBH) detection (T1 hypointensity) is a visual assessment showing some variability and therefore a need of larger sample sizes to notice a reduction over a short period of time. The predictive value of these measures in individuals is still very unclear.

GIOVANNONI, London, UK: CSF and other biomarkers. How to use them?

Neutralizing antibodies (NAB's) are probably the most validated soluble biomarker for therapeutic non-responsiveness in MS. Recent work has shown that the absence of induction of MxA gene expression by INFb in NAB+ patients with MS reflects a complete loss of bioactivity. The implication of this finding is that MxA is a potentially very good biomarker for non-responsiveness to IFN-treatment. MxA-induction has been used in a three years follow-up study of MS patients. This has shown that MxA-response is a predictive marker for relapse rate. In fact MxA proved to be much more predictive than MxA incorporated with NAB's (Malucchi, 2008). Next step is to incorporate MxA as a biomarker in the algorithm to decide when to stop treatment.

In terms of use of biomarkers in risk reduction, we are all aware of the problem of development of progressive multifocal leukoencephalopathy (PML) after natalizumab treatment. In fact, 24 cases have been reported since the drug was relaunched. In order to bring the natalizumab-levels down in PML cases as quickly as possible, a plasma exchange trial was performed with natalizumab plasma concentrations as a biomarker. Using 5 plasma exchanges it is possible to bring the natalizumab concentration down in 10 days to a level that is spontaneously reached only after 10-12 weeks. We do know now that plasma exchange is an effective method in PML.

Another example of the use of biomarkers with natalizumab is the assessment of JC virus (JVC) to determine the risk of developing PML. It is known now that this risk of developing PML is correlated to treatment duration and number of natalizumab infusions (particularly after 18 infusions). The seroprevalence of JCV (a type of human papillomavirus) among the UK population is around 50%. From PML case reports we know that PML does not occur if you do not have JCV. This feature has been incorporated in a newly designed natalizumab risk profiling and monitoring overview, in which patients without JCV have been placed in a low risk group for PML.

Soluble VCAM-1 is the best documented of all soluble markers. Patients with high levels of induced sVCAM-1 are much more likely to be in the responder category than patients with lower levels. Unfortunately, this is confounded by the presence of NAB's. If you have NAB's then you do not induce sVCAM-1.

Cellular markers are probably more useful for measuring therapeutic responsiveness in MS than soluble markers. CD56 NK cells is the best documented cellular marker. The literature is extensive in showing that MS patients have a deficiency of CD56 NK cells. Also a good correlation has been shown between levels of NK cells and MRI activity in a daclizumab trial. This was the first time that this marker was used in a clinical trial as a measure for therapeutic responsiveness.

Type 1 interferon signature in monocytes is an example of the use of systems biology (genomics, transcriptomics, proteomics, metabolomics, cellulomics) as a source for MS related biomarkers. This is a gene profile that predicts responsiveness to INFb.

Neurofilament (NFH) is an axonal damage marker. The level of neurofilaments in the CSF has been pretty well validated in a large number of animal models and is a very predictive marker for long-term prognosis in MS. It has been shown to correlate with disease progression in 3 years. If this biomarker holds up, we may be able to use CSF NFH levels to enrich neuroprotective trials (e.g. by profiling patients based on their NFH status), and also as a surrogate endpoint for treatment response.

Philosophical musings

Over the past decades the definition of (and criteria for) MS have been subject to change. Each time that the diagnostic criteria are adapted, the actual definition of MS changes. So by moving patients from Poser to McDonald criteria you actually improve the prognosis, both for MS and CIS patient groups. This is what is called the ‘Will Rogers phenomenon' in MS. A similar development has happened in NMO.

So when new MRI- or biomarkers are developed and become part of the diagnostic criteria, this is changing the definition and the profile (including the prognosis) of the disease. This argues for the advisors in the criteria commissions to hold back in changing the criteria.

In conclusion, we have biomarkers (becoming) available for measuring therapeutic (non-)responsiveness and risk reduction. Their use depends on the willingness of the pharmaceutical industry to add these markers to their trials. The real value of biomarkers needs to be established in line with the development of new drugs, taking them through the consecutive phases of clinical development.

COMPSTON, Cambridge, UK: The genetics of multiple sclerosis: susceptibility, course and pharmacogenomics

The interesting question to be asked in the genetics of MS is whether the factors which confer susceptibility will also be those that influence response to drug treatment, or whether those genetic effects, if there are any, are going to be specifically different. Expectations are that the latter of the two will be the case.

Over the years there have been many groups which have contributed to cataloguing factors that contribute to the genetics of MS. This is also something that the International Multiple Sclerosis Genetics Consortium (IMSGC, http://www.imsgc.org/) has been involved in over the past years.

The experiment that the IMSGC has done in terms of susceptibility was actually quite simple and rather unimaginative. In a case controlled association study they pooled the genetic data of large numbers of MS patients and compared those to the data of healthy controls in order to detect possible genetic markers for the disease. Three genome wide associations screens have been carried out thus far. The first screen showed the distribution of 500,000 single-nucleotide polymorphisms (SNP's) as possible markers, typed in some thousands of individuals and controls. The second screen (also published) was controlled by the Wellcome Trust Case Control Consortium (WTCCC). This screen was looking at 15,000 nonsynonymous SNP's (i.e. enriched for identifying genes) for genes involved in MS susceptibility. A third screen is the second screen controlled by the WTCCC. This IMSGC-WTCCC2 study is being analyzed, and is not yet published (expected early 2010). It is a large study involving 10,500 MS cases and 19,000 controls. The results from this screen are meta-analysed (GWAS) together with the data from the previous two screens by early 2010, involving 16,000 MS cases and 30,000 controls.

Essentially, this experiment is aimed at detecting suggestive genetic associations and by adding higher numbers of cases and controls shift the p-values from levels of suggestive or highly suggestive to levels of high probability. By doing so the p-values of the first identified genes IL-2RA and IL-7RA shifted from 10^-7 to 10^-20 or even lower. As a result some 25 new genetic defects were added to the list in the last 18 months after three decades of complete silence. An example is the CD58 marker that has moved from suggestive to quite strongly statistical significance. Also some genes have been identified that were shared between MS and other auto-immune disease. IL-2RA was found to be linked to MS as well as type 1 diabetes. In fact different alleles of IL-2RA contribute to susceptibility and resistance in MS and diabetes.

What is striking about the growing list of MS related genes is that the vast majority of the newly identified factors are the ones that are directly involved in T-cell and B-cell pathways, and particularly also in the signal transduction pathways operating in the immune system. And very few of these genes, as yet, do not have an immune function. Some have a specific neurobiological function and others have a completely unknown function. So the next step is to explore pathways that are suggested by these clusters of factors that have been identified, e.g. the IL-7 pathway. By looking at the IL-7 pathway two new factors could be added to the list.

With regard to the genetic influence on the clinical course of MS, all that can be said comes from the classic families analysis. The GAMES analysis looked to see if there was concordance with clinical severity course within families in which more than 1 pedigree had MS. The tendency to progress appeared to be a familiar feature, either with progress from onset or to progress after a period of relapsing-remitting disease. So the way of progression may be genetically driven, but we know nothing about the details.

Of all the genes that were found to confer to susceptibility, only one has emerged from any kind of pharmacogenomic analysis. And that gene was GCP5 (Glycipan 5 precursor), the top hit in the GeneMSA screen. Further investigation into this gene conferred some susceptibility to treatment response with INFb. Other possible markers for response to INFb treatment are still under investigation.

An annoying AE of treatment with alemtuzumab is that 20-25% of patients develop a secondary autoimmune disease. It has been found that high levels of IL-21 pre-treatment is a strong biomarker for this complication and that high levels of IL-7 pre-treatment are protective. High levels of IL-21 and IL-7 are strongly correlated with genotype for high producers of IL-21 and IL-7, which are two genes that have nothing to do with susceptibility of MS.

In conclusion, the heterogeneity of MS is very complicated. The IMSGC-WTCCC2 study will be published in 2010 and will add 25 genes to the list to total 30-40 genes that are related to MS susceptibility. So there is clearly a very complex genetic background apart from environmental links. One of the things that genetics has told us so far, is that there is this fundamental inflammatory conduit by which this disease is driven. This is confirmed by the fact that the new 25 genes involved in MS are mainly directly involved in T-cell and B-cell pathways and immune signalling. That is not to say that there won't be some additional factors which also drive susceptibility, and may also be involved in shaping the course of the disease (either in a good or bad way). And indeed there will be genes that do influence treatment, although these will most probably be secondary (and not directly part) of the set that is driving susceptibility.

SELMAJ, Lodz, Poland: Arguments for ending treatment – when and how?

There is a wise proverb that says that it does not matter how you start, it is important how you finish. Arguments to stop treatment can be categorized into three blocks. The first argument to stop treatment are (serious) adverse events. Relevant to this topic is benefit versus risk assessment and lack of patient compliance. A second set of arguments is linked to lack of efficacy. A related topic is how to deal with patients who progress from RRMS to SPMS. A third set of arguments is related to patient that have responded to therapy. Should we consider some kind of treatment vacation for them or should we even be optimistic enough to believe that treatment has induced long lasting remission and therefore discontinue treatment?

DMD treatment can cause a long list of side effects, some of which can really influence patients' decisions to discontinue treatment. The occurrence of side effects is a main reason for discontinuation of treatment: 71% in case of INFb treatment and 45% in case of GA treatment. Between 17 and 39% of patients have discontinued their DMD treatment for MS after a period of 3-4 years. On mitoxantrone the discontinuation rate is much lower (5% in 2 years).

There are several definitions from various groups about DMD treatment failure. The EMEA definition mentions ≥ 1 relapse during one year of treatment and/or ≥ 9 new lesions in T2-weighted image or >1 Gd+ lesion during one year of treatment. The Multiple Sclerosis Therapy Consensus Group (MSTCG) provides a more flexible definition with room for interpretation by the clinician. According to the MSTCG partial or full treatment failure is defined as the occurrence of further relapses within the first year, a confirmed EDSS progression as well as MRI parameters of ongoing or even increasing disease activity.

NAB's is a validated biomarker for therapeutic non-responsiveness in MS. Patients developing NAB's have a worse response to therapy (with IFNb), show a higher relapse rate and a faster disability progression. Also NAB's are a predictor for disease progression. Still, the interpretation of NAB+ as an argument to discontinue treatment is disputed. According to the AAN there is insufficient information on NAB testing regarding when to test, which test to use, how many tests are necessary and which cut off titre to apply. In case of lack of efficacy an escalating immunomodulatory strategy can be followed.

When reviewing the lack of efficacy of treatment, we also need to include the natural progression of MS. The transition from the relapsing remitting phase to the progressive phase of MS is marked by a lack of efficacy of immunomodulatory drugs due to diminishing influence of inflammation on disease progression.

There are several studies that have shown that an early start and continuous DMD treatment provides better long-term results in delay of disability progression (EDSS) than delayed start and/or episodic DMD treatment.

Nonetheless, the MSTCG argues that it appears to be acceptable to gradually discontinue DMD treatment at the expressed wish of a patient after at least three years of disease stability (no relapses, no clinical disease progression, stable MRI). A word of caution is advised when deciding to discontinue treatment, for cessation of INFb has shown that there is a considerable increase of annualized relapse rates and a decrease of time to 1^st relapse even within the first 6 months after discontinuation.

Atypical syndromes

MEINL, Munich, Germany: On humoral immunological aspects in the brain of multiple sclerosis patients

B-cells, plasma cells and antibodies have a long established role in MS. With intrathecal Ig production, oligoclonal bands, B-cells are present in the lesions and in the meninges. And there is also Ig and complement deposition in some lesions. The therapeutic success of plasmapheresis in some patients and the successful therapy with anti-CD20 indicates that the B-cells and antibodies are more than a biomarker and that they actively participate in the disease.

B-cells in inflammatory CNS diseases are not exclusively destructive. Many B-cells can be distinguished that show pro-inflammatory features and that may promote tissue destruction. B-cells are very efficient antigen presenting cells for their antigen that is recognized by their receptors, and they can produce pathogenic antibodies (localized inside or outside the CNS). On the other hand, B-cell can also be very beneficial for lesion development. They can produce IL-10 and neurotrophic factors and in some animal models Ig to oligodendrocytes might promote remyelination.

B-cell activating factor of the TNF family (BAFF) is one of the most potent humoral factors that promotes the survival of B-cells in the CNS. 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 is produced by astrocytes. In active MS lesions the expression of BAFF is as abundant as in lymphatic tissue (tonsils, adenoids).

Recently is was found the local IgG production in the MS lesion correlates with the local BAFF production.

This all adds to the view 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 modulating role of BAFF in the pathogenesis of MS is supported by a number of findings. BAFF is induced and linked to local Ig production in MS lesions. INFb-treatment has been seen to elevate BAFF, which is something that would not be expected from a beneficial MS treatment. A clinical trial has showed that atacicept (a soluble receptor for BAFF) worsened MS. And recently BAFF was reported to target the neuronal Nogo-receptor The inflamed CNS provides a survival niche for plasma cells. BAFF and its brother APRIL are an important factor for the survival of plasma cells, as are several chemokines (like CXCL12). This long-term survival niche is probably the reason why there are persisting oligoclonal bands in MS patients. These plasma cells are very tough and difficult to attack therapeutically for they do not express CD20 and would therefore escape rituximab therapy.

There are only a few known pathogenic antigens in MS that qualify as targets for auto-antibodies produced by plasma cells inside or outside the CNS. These targets are aquaporin-4 and myelin oligodendrocyte glycoprotein (MOG). MOG related antibodies have been investigated for nearly 20 years. Elevated MOG-levels have been reported in childhood demyelination in ADEM, but the data in adult MS are still controversial. This raises the question, whether there may also be other antigens involved in MS.

Key findings from our own research of anti-MOG antibodies add to the view that these antibodies are really pathogenic:

- The FACS cell based assay is clearly superior to ELISA or Western blot in depicting conformationally intact antibodies to MOG.

- These antibodies are present in 20-30% of childhood demyelination. These antibodies were seen in 78% of children aged below 5 years, but were absent or low in adult MS.

- The antibodies seem to disappear in the monophasic disease and to persist in childhood MS. The level of these antibodies at disease onset were not linked to persistence.

- The isotype of these antibodies was IgG1 (an complement activating isotype)

- Epitopes: there was overlap with pathogenic ones in animals

In search for the identification of novel targets for autoantibodies in MS we used a method of purifying human myelin glycoproteins. The idea behind this was that glycoproteins are typically expressed at the membrane surface and therefore accessible to antibody attack. By using western blots against known myelin components and mass spectrometry neurofascin and contactin-2 were identified as targets of autoantibodies in MS.

Neurofascin comes in two different isoforms: a myelin protein (NF-155) found in the paranodes and a neuronal protein (NF-186) concentrated at the nodes of Ranvier. Antibodies from MS patients recognize both NF-155 and NF-186. In animal studies these neurofascin antibodies are seen to target the nodes of Ranvier and induce an axonal injury. Axonal injury is linked to inflammation both in the acute and the chronic phase of MS. Auto-immunity to neurofascin is one of the several mechanisms that can induce an axonal injury.

Contactin-2 (also called transient axonal glycoprotein 1, TAG-1) is found in the juxtaparanodal region, both on the axon and on the myelin sides. It plays a role in axons during development but is also expressed in adults. A difference was found between patients and controls in the T-cell response to contactin-2. The MS patient group showed a higher level of T-cell proliferation to this antigen. In animal models T-cells specific for contactin-2 were seen to preferentially target the grey matter and induce inflammation.

FUJIHARA, Sendai, Japan: Neuromyelitis optica and astrocytic damage in its pathogenesis

NMO is a disorder consisting of fulminant bilateral optic neuritis (ON) and myelitis occurring in close temporal association. In 2004 NMO-IgG was reported (by Mayo Clinics in collaboration with Tohoku University) as an NMO-specific autoantibody. It was sensitive in 60-70% of cases with NMO, but also quite specific (90-100%), not only for typical NMO but also for partial NMO, e.g. recurrent ON. In 2005 aquaporin-4 water channel was identified as the target antigen of NMO-IgG. This is a water channel predominantly expressed in the CNS. In the CNS aquaporin-4 is mainly expressed on astrocytes (in the foot process and against the pial surface). The expression of aquaporin-4 is greater in the central grey matter than in the white matter. We have now collected more than 1,000 positive cases of aquaporin-4 antibodies in our database. From an analysis of 303 of these cases the following observations were made: 90% of cases are females, the average age of onset is 39 years (with a range of 3-86 years), mean EDSS is 5.5 (meaning that most patients need assistance in walking), 31% are blind in at least one eye (no light perception), only 13% have oligoclonal IgG bands (unlike MS) and 89% have long spinal cord lesions (longer than 3 vertebrae segments). So the absence of long spinal cord lesions doesn't rule out the possibility of NMO. Interestingly, 43 patients received INFb. In 11 cases treatment was discontinued due to side effect, in 25 cases exacerbations or no effect occurred and in 7 cases relapse rate remained unchanged. All adding up to 43 cases (100%) of non-response to IFNb treatment, which is quite different from the efficacy of IFNb as a first line DMT treatment in MS, suggesting that NMO is different from MS.

NMO patients suffer from severe disability in vision and motor function. In previous studies EDSS was used to rate those functional disabilities. However, EDSS is not designed to depict all the clinical problems of NMO because EDSS does not reflect some of the frequently occurring NMO symptoms such as severe pain and painful paraesthesias. When using quality of life measurement (with SF-36) in NMO, MS and controls, NMO scores were significantly lower than MS scores in physical function, bodily pain and general health. Bodily pain is caused by a unique pathology of NMO and is often the major complaint of NMO-patients. Unfortunately, there is no drug therapy that successfully relieves patients of their pain. In order to improve quality of life in NMO, the first thing to do is to decrease relapses in order to prevent visual and motor disability, but for the remaining neurological symptoms the development of symptomatic therapies against pain is the most important.

Sometimes patients with clear NMO features are found to have seronegative aquaporin-4 antibodies. In these patients aquaporin-4 antibodies may have become undetectable after immunosuppressive therapy with corticosteroids. There are some differences in spinal cord lesions between the chronic stage of NMO and MS. In NMO spinal cord lesions are usually longer than 3 vertebrae segments, whereas a majority of MS spinal cord lesions are shorter than 1 vertebrae segment. In addition, spinal cord lesions in NMO are localized in the central grey matter, whereas MS spinal cord lesions are distributed in the lateral and dorsal white matter parts. In the aquaporin-4 autoimmune syndrome patients show ON, NMO or myelitis, but they may also show only brain lesions. For instance, extensive white matter lesions with symptoms of disturbed consciousness may point towards NMO. In some patients these lesions may be presented as the onset symptom without previous history of NMO. Treatment has to be started directly in order to prevent blindness or motor disability occurring in the following NMO relapse.

The presence of astrocytic damage probably distinguishes NMO from MS, which instead is characterized by demyelination and gliosis. There are three lines of evidence that support that astrocytic damage is a key feature of NMO:

- There is extensive loss of aquaporin-4 and glial fibrillary acidic protein (GFAP, an astrocytic proterin) in NMO lesions pointing towards degeneration and loss of astrocytes.

- In NMO relapse there is remarkable elevation of CSF-GFAP levels (by factor 10,000), whereas levels in MS relapse are comparable to controls. CSF-GFAP levels have been found to be significantly correlated to EDSS a spinal cord lesion length in NMO.

- Aquaporin-4 antibodies have been shown to be pathogenic both in vitro and in vivo.

In summary, in NMO aquaporin-4 antibodies have diagnostic, therapeutic and pathogenic implications. Atrocytopathy mediated by aquaporin-4 antibodies distinguishes NMO from MS.

MONTALBAN, Barcelona, Spain: Susac's Syndrome, its diagnosis and treatment

In 1979, Susac and colleagues described the cases of two young women who had paranoid psychosis, hearing loss, progressive neurological dysfunction and multiple branch retinal artery occlusions. Six years earlier, in 1973, two similar cases were described unpublished, and were interpreted as seronegative lupus. Since it was described until now about 100 cases have been reported, although the real number of patients with Susac's syndrome (SS) is probably much higher. SS is a rare syndrome, but the frequency is probably underestimated. It affects mainly young women, but men can also be affected. The mean age at onset is 30 years.

The etiopathogenesis of SS is unknown. It is supposed to be an autoimmune disorder with an immune-mediated endotheliopathy which affects the microvasculature of the brain, retina and the inner ear. Perhaps anti-endothelial cell antibodies may play a role. In the brain some small areas of necrosis (microinfarcts) can be found within the cerebral cortex and white matter and also some neuronal, axonal and myelin loss

The clinical features show a classical triad of encephalopathy, branch retinal artery occlusions (BRAO) and hearing loss. The brain involvement is characterized by headache, which is quite common. It can appear several months before encephalopathy and is sometimes accompanied by visual disturbances (differential diagnosis migraine with aura). Encephalopathy is frequently associated with acute psychiatric disorders (APD), such as paranoid behaviour and hallucinations. Cognitive symptoms are frequent also: short-term memory loss, apathy and disorientation in time and space. Seizures are common and other neurological signs (such as pyramidal and cerebellar signs) can occur.

Retinal involvement causes symptoms such as visual field defects, photopsia and ‘black spots'. Inner ear involvement includes hearing loss which is quite frequent also. Lower tones are the first to be affected. It is generally subacute but can also be acute. Bilateral involvement is frequent although it is normally asymmetrical. Tinnitus and vertigo are quite frequent also. It is not mandatory that all three locations are clinically eloquent. People can experience a combination of the three sets of symptoms, which can appear together at the same time or in different time points. Permanent deficits may occur, which may be due to ischemic changes. This could be really serious.

On MRI white matter lesions are frequently seen, mainly in the corpus callosum (appearing as snowball like lesions) but also in other locations. Deep grey matter involvement and Gd+ grey and white matter lesions are frequent (70%).

There is no consensus of treatment of SS. Treatment should be prompt, aggressive and sustained, to avoid the dreaded residuals of dementia, deafness and blindness. In the acute phase high dose steroids (IV methylprednisolone: 1g/day x 3 days) are in order. Also cyclophosphamide (7 monthly pulses) can be very useful.

IVIG (2g/kg/day x 7days/month, x 6 months), mycophenolate mofetil (after CTX, daily dose, as ‘maintenance'), rituximab (not yet been used for SS!) and plasmapheresis have been described.

In quite a number of cases SS has been confused with MS and we have to be aware of that.

PALACE, Oxford, UK: Acute inflammatory white matter disorders

When looking into acute inflammatory white matter disorders that are not MS there are many conditions that could be listed under this title. These include ADEM (acute disseminated encephalomyelitis), ATM (acute transverse myelitis), monophasic Devic's disease and aquaporin-4 antibody disease (NMO), and CIS that does not develop into MS. This presentation focuses on ADEM.

ADEM was first noted after vaccination with smallpox and was also correlated with other vaccinations (1:600). This was caused by animal CNS proteins in the vaccine producing EAE. Nowadays, post vaccination ADEM is quite rare (0.1-0.2/10⁵ vaccinated individuals). Post-infection ADEM is much more common, with 75% of ADEM patients having a medical history of prior infection. 1 in 1,000 cases of measles are known to develop ADEM, which might be a direct infective neurotropic effect. Other known precipitating factors have been reported with drugs (sulfasalazine, methotrexate, anti-TNFa MAB, herbal remedies) and after snake bite.

There is a lack of validated diagnostic criteria in ADEM. ADEM usually occurs in children and has an incidence rate (0.4-0.8/100,000 per year) which is about 1/10^th of that of MS. Variations in case series of ADEM are based on inclusion differences based on requirement of a precipitating cause, associated encephalopathy, presence of MRI abnormalities or lack of subsequent relapse (monophasic). In pathology there is clearly overlap between ADEM and MS but there are also some noted differences. One of the most distinct differences is that demyelination in MS is confluent, whereas in ADEM it forms a perivenous sleeve.

With regard to clinical features of ADEM and MS, the following differences are noted:

- ADEM is monophasic, MS is chronic

- Gender ratio (F/M): ADEM 1:1 / MS 2⁺:1)

- Age: ADEM affects mostly children / MS affects young adults

- History of previous infection: 75% in ADEM / 38% in MS and relapse associated

- Prodromal illness often occurs in ADEM (headache, fever, general malaise) and not in MS

- Seizures are seen occasionally (17%) in ADEM children, and are rare in MS

- Encephalopathy is common in ADEM, not in MS

- Onset to the most severe disability is much shorter in ADEM (4-5 days) than in MS (10-14 days). The onset in ADEM is very classically poly-symptomatic versus mono-symptomatic in MS.

- ON in ADEM is classically bilateral versus unilateral in MS. Also the visual loss in ADEM is much worse.

- In ADEM there is higher serum pleocytosis and in the CSF less OCB's, higher protein and lymphocytes levels than in MS

- On MRI there is more resolution in ADEM, there are new lesions only in MS and lesions of different age are seen only in MS. At onset MS patients usually have a lower lesion load, whereas in ADEM patients the lesion load is usually quite extensive and sometimes with haemorrhage. Lesion appearance in MS is more well defined versus confluent or diffuse in ADEM. Periventricular sparing is more classic for ADEM than MS and MS lesions are commonly located near the corpus callosum.

There are two task forces that have taken up defining criteria for the diagnosis of ADEM: a pediatric task force (International Pediatric MS Study) and the US National MS Society Task Force. The diagnostic criteria are that clinical onset has to be polysymptomatic and there has to be encephalopathy, either with altered consciousness, behavioural or cognitive changes. The disease evolution may take 1 week up to 3 months with at least some improvement or recovery. The required MRI features are not strict. MRI should predominantly show two features that are symptomatic and that should be acute. Also some distinct criteria for relapse have been defined in order to differentiate ADEM from MS. There is still controversy about the differences between multiphasic ADEM (MDEM) and MS in terms of nature and location of new lesions. Basically the question is whether MDEM isn't the same as MS.

With regard to treatment of ADEM there is not much evidence available (class III evidence). Standard therapy is high dose IV methylprednisolone, tapering with a longer course of oral steroids (3-6 weeks) than would be normally done with MS because patients tend to rebound. Non-responders are given plasma-exchange or intravenous immunoglobulin. In aggressive forms of ADEM mitoxantrone and cyclophosphamide have been suggested.

The majority of patients (50-75%) recover fully and 70-90% show major recovery. There is a mortality rate of 5%. Poor recovery is associated with hyperacute onset and very severe disability at onset. There are variants of ADEM with haemorrhage or necrosis, representing about 2% of ADEM cases in childhood. These are rapidly progressive forms associated with haemorrhage or necrosis that need to be treated more aggressively.

In trying to classify MS and related conditions, the following model may be suggested.

The point of trying to distinguish between these conditions is that there iare major differences between many of these monophasic conditions. And the big questions is what are the mechanisms that render the monophasic inflammatory CNS conditions self-limiting? And maybe this provides us with clues how to manage MS.

HEMMER, Munich, Germany: Auto-antibodies to CNS proteins in children with first demyelinating event – implications for diagnosis and prognosis

Pathogenesis of MS includes a predisposition to genes, most of which are immune genes related to the activation of T- and B-cells. Also the environment plays a role in pathogenesis, but we still know very little about the factors that drive the disease. Both genes and environment have an impact on the immune system, probably predominantly on the acquired immune system. There are good arguments that suggest that the early phase of MS is mainly driven by the immune system. However, the target antigens of the acquired immune response are still unknown, but given the success in the research of NMO where aquaporin-4 was identified as an important target, there is still hope that there are specific antigens in MS that are driving the immune response. It is most likely that these antigens will be identified from antibody experiments and probably we have to look very early in the disease course.

When working with antibodies a number of issues have to taken into account. If an antibody causes or adds to the disease in MS, it is likely that the antibody recognizes an antigen that is expressed in the CNS. Traditionally we have looked at the myelin sheath and at oligodendrocytes, but probably we also have to look at other targets such as astrocytes and microglia

Also, if an antibody causes or adds to the disease in MS, the antibody has to recognize certain targets. There are three possible scenario's how an antibody could cause damage: by inactivation of extracellular proteins (e.g. neurotrophin), by receptor blockade or (most probable) by cytotoxicity in which the antibody binds to a membrane protein expressed on a particular CNS cell which, via the activation of the complement cascade or by activation of NK-cells or macrophages, leads to an attack against this expressed cell and eventually to the death of this cell. So the important message is, that if you want to look for targets that are relevant to auto-immune diseases in the CNS, you probably have to look at membrane proteins and extracellular proteins. Of these the membrane proteins are the favourites.

Antibodies that cause harm need to recognize antigens in their real native form, which means that they have the right conformation. Traditionally we have looked for antibody response in many autoimmune diseases by using assays that are based on key nature proteins. So when doing an assay, the protein is denatured and thus from a very complex protein a linear aminoacid sequence is made and analyzed for antibody response. It has been well established from animal studies, that if you have an autoimmune response in the CNS and you create antibodies to conformational and linear epitopes, then only the ones that recognize the conformational protein are relevant and can mediate damage. If you want to identify those antibodies that can harm, you need the right assays that display the protein in a way that is close to the way that it is expressed in the CNS.

Several experiments have been done with respect to antibody response to MOG. There are a number of reasons why MOG could be an interesting target in CNS autoimmunity. It is expressed in the oligodendrocytes, it is a membrane protein, it is expressed on the outer surface of the myelin sheath (so well accessible to antibody responses), it is encephalitogenic in different species (monkeys, rats, mice) and it induces pathogenic antibodies in EAE models.

High antibody titres to native MOG are found in up to 40% of children with a first demyelinating event (diagnosed with CIS or ADEM). Such antibody titers are not seen in pediatric controls. These titres are age dependent and seem to decrease over time.

Antibodies do not predict conversion to MS and do not discriminate between ADEM, CIS or MS. However they are not found in other neurological diseases. Antibodies induce ADCC in vitro, bind to rat myelin and seem to induce demyelination in inflammatory lesions in rat EAE. Anti nMOG antibodies occur in children, who are not infected with EBV (and have no immune response against EBV). This makes it unlikely that the anti nMOG antibodies response is solely driven by EBV.

The future

WOLINSKY, Houston, USA: Are there emerging treatments that promise to reduce progression of the disease?

There are a number of endpoints to be used in clinical trials. Only the clinical endpoint are the ones that are really important for registration and in general they are also the only ones that are really interesting for patients. When talking about progression, the usual relapse related endpoints (such as relapses per treatment arm, annualized relapse rate, time to first relapse, relapse severity) are relevant. But when talking about disability related endpoint, we should discuss how we can best measure accumulative disability and whether or not that meets a definition of progression. And then we can talk about phase related changes and see if a phase shift is being delayed.

One phase shift hasn't been attacked until now, and that is the phase shift of RRMS to SPMS. Now if we would do something there, this would really contribute to a phase shift reduction of some importance.

There are two possible trials that could be used to measure the prevention of disease progression in later disease. The first is a trial (type 1) to measure the impact of treatment on the transition to progressive disease (i.e. a phase shift reduction from RRMS to SPMS). And the second trial (type 2) would be to measure the impact of treatment on accumulated disability in progressive disease (once SPMS has been reached).

In considering a suitable design for these trials, a number of issues come up that need to be addressed. First it is a matter of clearly defining the subsequent disease phases (RRMS and SPMS). The current definition of SPMS mentions that SPMS is a long-term outcome of RRMS in that most SPMS patients initially begin with RRMS … and that once the baseline between relapses begins to progressively worsen, the patient has switched to SPMS. Now this is a definition that could clearly never be used in a trial designed to prevent progression to SPMS.

There are several operational definitions of progression that have been used. Perhaps it would be advisable to use the proportion of patients reaching EDSS 4.0 in these trials as a measure for prevention of progression to SPMS. From the natural history cohorts (like EDMUS) we can say that there is a certain amount of time needed for RRMS patients to reach EDSS 4.0. But once they reach EDSS 4.0 the remaining course of disease is more consistent and the time to go from EDSS 4.0 to 6.0 is fairly predictable. The most important thing is that EDSS 4.0 is the critical point at which we could operationally say that the patients have reached SPMS.

Measuring accumulated disability by using the proportion of patients that reach a certain level of EDSS sustained for a particular amount of time is quite a time consuming method. From the MAGNIMS cohort it was therefore suggested to use a composite endpoint of EDSS measurement or timed 25-feet walk or 9-hole peg test change. In that case it took 2 years for a proportion of 63% of patients to reach the endpoint. Since this probably didn't include a requirement for sustained change, the suggested type 2 trial would need to run longer in order to show that patients not only get to that level, but also stay at that level or may even worsen. From previous studies in progressive MS (most of which failed) we know that substantial amounts of patients can be expected to progress over the course of 2-3 years.

The suggested way forward in designing the mentioned type 2 trial is to return to controlled trials of progressive MS. Dividing up the groups (i.c. PPMS and SPMS) wouldn't be necessary, because these groups are still more similar than that they are different. We may need active comparator studies to make these acceptable when looking at the required duration of the trials. As SPMS is hard to define and PPMS is not always what it is said to be, functional definitions have to be adopted. For the type 2 studies the entry EDSS must be ≥ 4.0, not attained by one or two acute attacks (i.e. exclude RRMS with stable fixed deficits). Also EDSS must be < 6.5 because ambulation is still going to be a critical endpoint. The age must probably be >34. We could either decide to take patients who are MRI active or take patients who are MRI inactive or stratify our outcomes by that. Prior therapy may not be relevant but needs be recorded. The outcome would be the development of EDSS 6.0 or 6.5 sustained for 3 months depending on entry EDSS. This means that the trial duration must be at least 5 years to get the critical endpoints. The trial drug needs to be able to penetrate the CNS. It could be monotherapy, but then the drug would need to have a dual action (something that fingolimod is supposed to have). If such a monotherapy is not available then we need to have systemically inactive neuroprotectives on an anti-inflammatory platform.

HEESEN, Hamburg, Germany: Patient autonomy in treatment choices

The question playing in the background of patient autonomy in treatment choices is who is to decide on MS treatments and to what extend: patients or physicians? One clear trend that comes from a study in which doctors and physicians were asked at which PML-risk they wanted to stop natalizumab treatment, is that patient tend to be more risky than physicians. On the other side there is no evidence that patients (or physicians) overestimate the benefits of an MS drug treatment. When looking at cultural differences in autonomy preferences there are some indications that Mediterranean populations tend to prefer a paternalistic approach by physicians and Northern-European populations a more autonomous approach. When CIS patients were asked whether they would want to be involved in decision making about undergoing diagnostic MRI, a clear majority (2/3^rd) responded in favour of informed consent or shared decision making.

The EBSIMS study (Evidence-Based Self-management In Multiple Sclerosis relapses) was a multicenter, randomized-controlled study on effects of an education programme on relapse management in MS. This study was published in 2008 in Multiple Sclerosis. The study hypothesis was that trained patients would show more autonomous management of relapses, including delaying or refraining from treatment and possible oral self-medication with an increased feeling of control. The primary endpoint was the number of resignments from treatment and the number of oral self-medications. 150 patients (RRMS, 2 relapses within previous 24 months and no major cognitive deficit) were followed-up for 2 years. Only 13 patients (6.5%) did not finish the study and were lost to follow-up. There were no significant differences between randomized groups (mean number of relapses in two years were comparable). The results showed highly significant differences on primary endpoints in favour of the intervention group (receiving brochure and training). Secondary endpoints showed a higher decisional autonomy and less visits or phone calls to physicians in the intervention group. No differences in quality of life (HAQUAMS) or disability were found between groups after 2 years. Trained patients were also found to experience 0.6 less relapses in 2 years, which is not likely due to reporting bias. This may be an indicator that the sense of control and increasead autonomy suppress disease activity and not only works on well-being.

PEPADIP (effectiveness of a Patient Education Program About Diagnosis, prognosIs and early treatment in Patients with multiple sclerosis) is a second study, which is still ongoing. This study addresses the impact of an evidence based patient education programme (brochure and 4h group session) on patients with CIS and MS (<2 years) as compared to a stress and coping intervention for MS patients (so two active groups with different concepts of intervention). It is expected that the educational programme will increase relevant disease-related risk-knowledge and promotes informed choice. Also it is expected to promotes a sense of control, decision autonomy and satisfaction and as a result leads to an altered i.e. more rational approach to immunotherapies by reducing anxiety and depression.

In the area of therapeutic decision making we need strategies (preferably standardized) how scientific evidence is best communicated to patients. In parallel patients need to made clear about their personal values and anxieties. This has to be integrated with the experiences from physicians, patients and others. This has to be achieved in the background of ambivalence of physicians as healers on the one hand and service providers on the other.

In conclusion, patients need to be involved and also want to be involved in therapeutic decision making. There is evidence that patient information is feasible and has a positive impact on health behaviour. More research is needed on autonomy preferences, determining factors (psychological, medico-cultural) relevant to decision making and on the effect of behavioural interventions.

LIST OF ABBREVEATIONS

AAN American Academy of Neurology

ADEM Acute Demyelinating Encefalo-Myelitis

ADCC Antibody-Dependent Cell-mediated Cytotoxicity

AE Adverse Event

BAFF B-cell Activating Factor of the TNF-Family

BBB Blood Brain Barrier

BPF Brain Parenchyma Fraction

BRAO Branch Retinal Artery Occlusion

CC Corpus Callosum

CDMS Clinically Definite MS

CIS Clinically Isolated Syndrome

CNS Central Nervous System

CSF Cerebrospinal Fluid

DMD Disease Modifying Drug

DMT Disease Modifying Treatment

EAE Experimental allergic autoimmune encephalomyelitis

EBV Epstein-Barr Virus

EDSS Expanded Disability Status Scale

EDMUS European Database for Multiple Sclerosis

fMRI functional MRI

F/M Females / Males ratio

GA Glatiramer Acetate

Gd Gadolinium

GPRD General Practice Research Database

HLA Human Leukocyte Antigen

HTA Health Technology Assessment

HAQUAMS Hamburg Quality of Life Questionnaire Multiple Sclerosis

Ig Immunoglobulin

IL Interleukin

IFNb Interferon-beta

IFNg Interferon-gamma

ITP Idiopathic Trombocytopenic Purpura

IVMP Inta-Venous Methyl-Prednisolone

JVC JC Virus

GA Glatiramer Acetate

GFAP Glial Fibrillary Acidic Protein

KFS Kurtzke Functional Subscore

MBP Myelin Basic Protein

MOG Myelin Oligodendrocyte Glycoprotein

MRI Magnetic Resonance Imaging

MTR Magnetization Transfer Ratio

MS Multiple Sclerosis

MSC Mesenchymal Stem Cell

MSTCG Multiple Sclerosis Therapy Consensus Group

NAB Neutralizing Antibody

NAWM Normal Appearing White Matter

NEJM New England Journal of Medicine

NK-cell Natural Killer cell

NMO Neuromyelitis Optica / Devic's Disease

NNT Number Needed to Treat

NPC Neuro Precursor Cell

Nrf2 NF-E2-related factor

NSF Nephrogenic Systemic Fibrosis

OCT Optical Coherence Tomography

ON Optic Neuritis

OPC Oligodendrocyte Progenitor/precursor Cell

PASAT Paced Auditory Serial Addition Test

PBMC Peripheral Blood Mononuclear Cell

PML Progressive Multifocal Leukoencephalopathy

PPMS Primary Progressive MS

RCT Randomized Clinical Trial

RNA Ribonucleic Acid

RNFL Retinal Nerve Fibre Layer

RR Relative Risk

RRMS Relapsing Remitting MS

SAE Serious Adverse Event

S1P Sphingosine 1-Phosphate

SJL Swiss Jim Lambert

SLE Systemic Lupus Erythematodes

SPMS Secondary Progressive MS

TAG Transient Axonal Glycoprotein

TNF Tumour nNcrosis Factor

TGFa Transforming Growth Factor alpha

TGFb Transforming Growth Factor beta

VCAM Vascular Cell Adhesion Molecule

WM White Matter

--- END ---