25. Aetiology and origins of multiple sclerosis

Abstracts / Journal of Clinical Neuroscience 17 (2010) 1610–1638

23. Longitudinal employment change in multiple sclerosis and the importance of symptom management Elizabeth A. McDonald, Rex Simmons MS Australia People with multiple sclerosis (MS) tend to have lower participation in paid employment compared to those with other chronic diseases. Using the Australian MS Longitudinal Study, two largesample, self-report surveys of people with MS were performed four years apart, in which employment rates were measured, cross-sectionally and longitudinally. The reasons for employment loss and perceived risk of future employment loss were also assessed. A total of 1135 Australians with MS responded to the first survey, 1329 to the second, and 667 to both. Longitudinal loss of employment was 5.4% over the four years 2003–2007, a period of relative economic prosperity in Australia. By 2007, 56% of the MS cohort had lost employment due to MS and 64% were not in the paid labour force. Cross-sectionally, the age- and sex-standardised employment rates were significantly below those of the general population. Regression analysis indicated men were more likely than women to leave their employment because of MS, and older people were more likely than younger people. Level of occupational skill using Australian Bureau of Statistics (ABS) categorisation was not predictive of maintaining or losing employment. The main reasons reported by people with MS for employment loss involved ineffective management of symptoms of MS in the workplace, rather than workplace-related factors. These findings imply symptom management in an employment context needs to be addressed early in MS. Reference: Simmons RD, Tribe KL, McDonald EA. Living with multiple sclerosis: longitudinal changes in employment and the importance of symptom management. J Neurol, in press. doi:10.1016/j.jocn.2010.07.024

24. Use of the addenbrooke’s cognitive examination in multiple sclerosis Cullen M. O’Gorman a, Susan Freeman b, Simon A. Broadley a a b

Griffith University, School of Medicine and Gold Coast Hospital, QLD Griffith University, School of Medicine, QLD

Objective: To establish clinical utility of the Addenbrooke’s Cognitive Examination (ACE) in assessing cognition in patients with multiple sclerosis. We have compared the ACE to the MMSE, and explored relationships between ACE score, EDSS, lesion load on MRI and atrophy. Methods: Subjects were 83 patients with multiple sclerosis seen at the Gold Coast MS Clinic. ACE and MMSE assessments were undertaken. Demographic and EDSS data were recorded. The total parenchymal lesion load and intercaudate ratio (ICR) calculated from contemporaneous MRI brain images. Results: 14 patients demonstrated cognitive impairment on the ACE (scoring <86/100), with 8 patients detected by the MMSE (scoring <27/30). A correlation between low ACE score and greater disability was seen. ACE score also correlated with lesion load and ICR. Discussion: The ACE is a more sensitive screening test than the MMSE in our study. The MMSE has been shown to be insensitive in MS patients but due to familiarity and ease of administration it remains a popular screening tool. The ACE has advantages compared to formal psychometric batteries being quick to administer and also being free. Subcortical atrophy in MS can be approximated by linear measures such as the ICR. Prior studies showed an increased ICR in association with cognitive dysfunction, EDSS and disease duration.

1617

Our findings suggest that the ACE can detect the cognitive decline associated with subcortical atrophy. We also find association between higher EDSS scores and measured subcortical atrophy. doi:10.1016/j.jocn.2010.07.025

25. Aetiology and origins of multiple sclerosis Alastair Compston University of Cambridge, UK Despite early descriptions of familial multiple sclerosis, not until the mid-1980s was it finally established, based on recurrence risks in many categories of relatives of index cases, that there must be a genetic contribution to disease susceptibility. Following description of the HLA-DR15 association in the early 1970s, little further progress was made in identifying other susceptibility loci for over three decades. In the early 1990s, reagents emerging from the human genome project allowed systematic genome screening to be performed and these were quickly used in whole genome screens for linkage and association. However, these early efforts over-estimated the effect sizes of any additional susceptibility loci and were under-powered. Only with the formation of consortia (Genetic Analysis of Multiple sclerosis in EuropeanS [GAMES], 1999; the International Multiple Sclerosis Genetics Consortium [IMSGC], 2003; and the Wellcome Trust Case Control Consortium, phases 1 and 2, from 2005; ANZGene, from 2005; and others) was there a sufficient increase in the scale of these genetic analyses that led, from 2007, to the results that were based on studies combining adequate sample size with high density coverage of the genome. These experiments involved many billions of genotypes and presented new challenges for statistical evaluation of the results. Taken together, in 2010, the convergence of results across these screens makes it possible to list several genetic loci that appear to confer susceptibility to multiple sclerosis: these are HLA-DRB1, p = 8.9  10-225; HLA[class 1]C, p = 3.3  10-5; cd25 (IL2-Ra), p = 9.6  10-29; IL7R (IL7Ra), p = 5.5  10-20; TYK2 (Tyk2), p = 2.7  10-6; cd226 (DNAM1), p = 5.4  10-8; CLEC16A, p = 1.6  10-15; LFA3 (CD58), p = 3.10  10-10; TNFRSF1A, p = 1.59  10-11; ICSBP1/IRF8, p = 3.73  10-9; CD6, p = 3.79  10-9; CXCR4, p = 3.34  10-7; CYP27B1 [or METTL1] p = 5.4  10-11; and CD40 p = 1.2  10-7. These databases have also been used to validate, and in some case amend, reports (often attractive and mechanistically plausible as in the case of KIF1B) that could not subsequently be confirmed. The Wellcome Trust Case Control Consortium (WTCCC) has recently completed a substantially more comprehensive genome wide association screen using an independent set of more than 10,000 cases collected through the IMSGC and involving research groups from Australia, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Norway, Poland, Spain, Sweden, the UK and the USA. Samples were genotyped using the Illumina 670 Quad chip and analysed after appropriate quality control filters had been applied. Data available from almost 6000 UK subjects already generated as part of WTCCC2 were supplemented with those from other published studies to provide a sample of c12000 controls. This screen was powered to identify common risk alleles with an odds ratio of P1.2. The results consolidate and extend the recent exponential increase in knowledge on the genetics of multiple sclerosis. These were first presented at the American Academy of Neurology (April, 2010: Toronto). As now expected, the relative risk for each new contributing factor is small indicating that, alone, these genes make a small contribution to susceptibility; but the role of epistatic interactions and the extent to which the individual effects converge on coherent pathways remain to be determined. Replication is being performed

1618

Abstracts / Journal of Clinical Neuroscience 17 (2010) 1610–1638

using an additional cohort of 9000 cases; and meta-analysis of all available data relating to multiple sclerosis will follow later in 2010. The first wave of novel genes now shown to confer susceptibility to multiple sclerosis has mainly revealed products involved in pathways of the immune response, thus providing a further clue to the dominant disease mechanism in multiple sclerosis, and allows speculation on the origins of susceptibility to multiple sclerosis. Since the cases studied are not stratified for phenotype or other laboratory biomarkers, these data add weight to the interpretation that, although different effector mechanisms may be involved, the disease is complex but not heterogenous and driven by the core process of focal brain inflammation. doi:10.1016/j.jocn.2010.07.026

the assumption that common diseases such as PD are caused by common genetic variants. An alternative theory that seems applicable to PD, is that a number of rare variants of intermediate penetrance may be responsible. On the other hand, there is a suspicion that the greatest molecular risk for PD relates to a-synuclein expression and therefore these variants will exert their influence by modulating dose, handling or clearance of a-synuclein by the neurones. It is also plausible that these contributing common genetic variants may be located in existing PARK loci. doi:10.1016/j.jocn.2010.07.028

28. Conduction failure in the peripheral nervous system John Pollard University of Sydney, NSW

26. Epilepsy Terence J. O’Brien Department of Medicine, University of Melbourne, VIC The outcomes of treatments for epilepsy, both medical and surgical, are highly variable and largely unpredictable for an individual patient; for seizure control, adverse effects and psychosocial outcomes. Patients not infrequently need to be tried on a series of different treatments, in ‘‘trial and error” fashion, before the optimal choice for that individual is found. This is a highly unsatisfactory situation for multiple medical, safety, psychosocial and health economics reasons. The emerging field of Pharmacogenomics offers the promise to enable the vision of ‘‘Personalised Medicine” with the safer and more effective drug treatment. While clinically important advances have been made on identifying genetic markers for serious cutaneous adverse drug reactions, numerous attempts over more than a decade to identify markers that are predictive of seizure control have not been strong enough to be clinically useful in an individual and have failed to stand up to attempted replication. Most of the published studies have attempted to identify single nucleotide polymorphisms (SNPs) predictive for treatment outcomes. However for complex diseases such as epilepsy, affecting heterogeneous populations, it is unlikely that a single SNP will adequately explain treatment outcomes, which are likely to have multi-factorial determinants. Furthermore, gene-gene interactions make the discovery process more complex than for standard Mendelian traits controlled by a single locus, and non-genetic factors, and gene-environment interactions, are almost certainly also important determinants. Therefore the development of accurate predictors for the response to epilepsy treatments needs to consider multiple genomic and non-genomic variables. In recent years, genotyping technology has rapidly advanced and it is now practical and affordable to obtain high resolution genome-wide data in an individual. However, it remains a great statistical and bioinformatic challenge to optimally utilize these large, rich datasets to derive predictors that are clinical useful. doi:10.1016/j.jocn.2010.07.027

27. Parkinsons disease Malcolm Horne Florey Neurosciences Institute and St Vincent’s Hospital, VIC At the time of writing there are 16 Park genes yet despite the results of recent Genome Wide Association studies we can only explain about 10% of PD cases. However GWA studies are based on

This lecture will focus on mechanisms of conduction failure highlighted from recent studies in inflammatory neuropathy (IDN). For many years conduction failure in these demyelinating neuropathies was thought to result from loss of a passive insulator (myelin) around axons. This view was reinforced by a focus of research on the animal model Experimental Autoimmune Neuritis (EAN) induced by immunisation with heterologous myelin. However the emerging view is that molecular adaptations of myelinating Schwann cells facilitate saltatory conduction and that conduction impairment results from disruptions of these cells and the contribution they make to the molecular specialisations of the axon particularly at the nodes of Ranvier. The EAN model also for many years restricted the attention of researchers seeking target antigens in IDN to considerations of myelin proteins and lipids. Recent studies in patients however emphasised the role of ganglioside antibodies and gangliosides are enriched in neural membranes including the axolemma particularly in nodal and paranodal regions. In experimental studies in the rat, we injected monoclonal antiganglioside antibodies intraneurally or passively transferred them into animals treated to induce blood nerve barrier leakiness. Anti Gd1a and Gm1 antibodies caused a reversible conduction failure, widening of the node and a dose dependent degeneration, but no demyelination. Susuki et al. (2007) established the probable mechanism of this conduction failure in their AMAN Rabbit Model. They showed that anti Gm1 antibodies cause a complement mediated disruption of sodium channel clusters by disturbing the paranodal and nodal axoglial junctions which stabilise the Nav clusters to the nodal cytoskeleton. Immunohistochemical studies showed that channel dispersion was accompanied by loss of the adhesion molecules responsible for binding Schwann cell microvilli and paranodal loops to the axolemma. Further evidence for the relevance of nodal changes was produced by Lonigro and Devaux (2009) who studied nodal changes in EAN rats. In animals immunised with PNS myelin (which induces a B and T cell response) Nav clusters at nodes were disrupted and two cell adhesion molecules essential for Nav channel clustering, Neurofascin and Gliomedin, were selectively affected prior to demyelination; this was associated with auto antibodies to these molecules. We have measured antibodies to Neurofascin in patients with IDN and shown elevated levels in subsets of patients with GBS, CIDP and Multifocal Motor Neuropathy. In other studies we injected antibodies to Neurofascin into rat sciatic nerve and shown a reversible conduction impairment similar to that obtained with antiganglioside antibodies. These findings suggest that nodal molecules such as those involved in nodal and paranodal specialisations may represent target antigens in IDN. Furthermore conduction failure in these conditions is not caused by loss of a passive insulator