1 Genetic Analysis of Multiple Sclerosis in EuropeanS Jesus College, Cambridge 3 – 4 April 2000 2 SESSION 1 The analysis of complex traits Chairman : Kees Lucas Alastair Compston opened the meeting saying how pleased he was at the enthusiastic response to GAMES. He acknowledged the UK and Northern Ireland Multiple Sclerosis Societies, the Wellcome Trust, the Dutch Multiple Sclerosis Society and PE Biosystems for supporting the meeting, and thanked the Cambridge team, especially Katrina Dedman and Julia Gray for organising the event. GAMES follows a tradition which began in 1984 with groundbreaking epidemiological studies. Following the 1989 meeting in Cambridge, DNA collection started in earnest. Initial efforts concentrated on biological and immunological candidate genes for which markers were available, but it soon became evident that an alternative approach systematically screening the whole genome would be needed. Four linkage based whole genome screens have been completed. No regions of linkage with genome-wide significance have yet been identified, although several regions of 1 2 3 25 interest have been found . Heterogeneity and the presence of epistatic effects were illustrated by segregating the results according to HLA haplotype4. It has become clear that the genes conferring susceptibility to MS individually have only a modest effect. A change of strategy from linkage to family based association studies on a large scale now seems necessary. The GAMES experiment should accelerate and could solve the problem of the search for MS susceptibility genes. Our hope is that in five years the current map will have been refined, some peaks will have disappeared from the list of potential candidates for the location of susceptibility genes, whilst others will undoubtedly have become more significant. Some of these regions may be ubiquitous and some domestic to certain populations. Hopefully someone will have spotted a relevant gene 3 John Todd described his experience in the genetic dissection of another complex disease, IDDM. One of the main difficulties is to conduct a study with sufficient power. The main requirements for such studies are that large sample sizes and high throughput automated technology 5 are needed 6. Nominal p-values of at least 5x10-8 are required before claiming positive results (genomewide significance)7. Further correction for multiple testing involved in genome screens is needed, especially when analyses are stratified. For genes with modest effects (s <2) the number of samples required to achieve genome-wide significance can realistically only be achieved if either case-control or family based association studies, rather than sib pair based linkage studies are performed. The power of case control studies can be increased by using more controls than cases (given that controls are generally easier to collect). These controls could be used for studies of several diseases. Background population allele frequencies can be estimated by looking at non-transmitted alleles from parents of simplex cases. An experiment comparing populations from the United Kingdom, Sardinia, Finland and the United States has demonstrated that linkage disequilibrium (LD) appears to be similar in each population, and rather more extensive than previously suggested. Microsatellites up to 1 centimorgan apart are often found to be in LD. Professor Todd argued that the identification of susceptibility genes would ultimately require “direct” association mapping, i.e. the testing of all known Single Nucleotide Polymorphisms (SNPs), which would necessarily include the causative ones, in very large case control populations. He described developments, which could potentially 8 9 10 make such an approach realistic , including the role of DNA pooling . He suggested that autoimmune loci may be shared between diseases and summarised the evidence for chromosome 18 and 2 (CTLA4). David Clayton discussed problems associated with the statistical analysis of 7 complex traits . The power to detect the relevant genes is limited by their modest genetic effects, as reflected by the small recurrence risks which they determine. An alternative to typing all available polymorphisms is to rely on linkage disequilibrium and type a smaller numbers of markers (“indirect association studies”). Demonstration of population association requires markers to be very close to the functional gene (or the marker actually to be the functional gene). In terms of power, case control studies are the method of choice when testing for association, although 4 the selection of controls is crucial since unobserved population stratification may create bias. However, this problem only has a substantial effect in extreme situations, when it should be easy to spot. Family based association studies (i.e. TDT) protect from effects of stratification, but are less powerful because of the intra- 11 familial genetic similarities . The effects of multiple testing have to be taken into account and require more rigorous significance levels to be employed. A linear correlation exists between log (significance) and required sample size. Thus, increasing significance levels to the high levels suggested by Risch should not be prohibitive. Linkage analysis does not depend on the vagaries of population history and, for many of the genes involved in complex traits, it may well be possible to demonstrate linkage with genome-wide significance if efforts are made to collect sufficient families. In short, it would be premature to abandon linkage and efforts with this method should continue. Arne Svejgaard described the relation of HLA and multiple sclerosis. The HLA complex was sequenced in 1996. It extends over 3.6 Mbp (3.6 cM) on the short arm of chromosome 6. It has 128 functional genes, 40% of which have immune function. The associations between multiple sclerosis and HLA A3, DR2 and B7 were 12 demonstrated in the 1970's . Spurkland et al. demonstrated in 1997 that susceptibility is conferred both by HLA DRB1*1501 and HLA DQB1*060213. The association with TNF results from linkage disequilibrium. Madsen et al. showed recently in an animal model that HLA-DR2 can mediate both induced and spontaneous disease resembling MS, by presenting an MBP self-peptide to T cells14. Narcolepsy, probably caused by alterations in the Orexin 2 receptor, is highly associated with the same HLA haplotype as multiple sclerosis15. General problems arising in the study of candidate genes were discussed by Jan Hillert. The strategy for suggesting the right candidate is to choose genes with specific biologic function and those mapping to an area of interest from previous genome screens or implicated in animal models16. Replicating associations is made difficult because of the genetic and allelic heterogeneity and the epistatic effects. Relevant polymorphisms may be old and linkage disequilibrium therefore only extending over relatively short distances. Positive findings should be confirmed with linkage and association in at least two separate populations. Candidate genes of interest in multiple sclerosis include CTLA417 18, which is important in T cell activation (negative signal). Apo E alleles (which are associated with sporadic and familial 5 19 Alzheimer's Disease) may affect clinical course and severity and the Interleukin 1 20 complex which may have a combined effect with HLA-DR on susceptibility to relapsing multiple sclerosis and disease progression. Future candidate gene studies should benefit from information emerging through sequencing the human and mouse/rat genomes. Bertrand Fontaine stressed that linkage and association based studies can be employed in the search for candidate genes, which may not only be chosen to look for disease susceptibility but also for treatment response or clinical course. Difficulties in candidate gene approach include the high number of potential candidates, and the possible absence of intragenic polymorphic markers within the region of interest; for example, the oligodendrocyte growth factor complex offers very promising candidate genes, but consists of more than 50 different factors, which have effects on oligodendrocyte migration, differentiation, regeneration, survival or proliferation, making it is difficult to select the most promising candidate. There may be a role for one of the oligodendrocyte growth factors (TGFb3) in interaction with HLA421. Future studies need to employ large sample sizes and consider clinical or para-clinical subgroups as well as stratification for HLA. The Cambridge group then talked about their vision of the GAMES experiment. Alastair Compston showed a map illustrating where the various groups expressing interest are based and emphasised that attending the meeting does not imply signing up to a commitment. Some groups have already formed consortia (i.e. the Nordic group), whilst others work independently. Several groups have significant experience in the genetics of multiple sclerosis. GAMES is a multi-centre linkage disequilibrium screen of the whole genome in several European populations. 6000 microsatellite markers have already been purchased and are available for those investigators wishing to play GAMES. The sequence will be a screen of the genome using 6000 markers in DNA pooled from at least 200 cases and 200 controls, followed by a second screen using family based samples (a minimum of 200 trio families pooled into two samples). This means, that four pools will be screened by each GAMES player (requiring a total of 24,000 genotypes). In addition to these association studies, efforts to collect multiplex families will continue, and some linkage genome screens will also be performed 6 within GAMES. There are likely to be several types of GAMES players. Type 1 players will use markers provided by Cambridge, but will collect and screen samples locally; those will be groups who already have access to fluorescence based sequencing machines using ABIs 373, 377, 310 and 3700. Type 2 players will send a member of the group to Cambridge to genotype and analyse samples which have been collected and pooled locally. Type 3 players will collect and pool samples locally and send these to Cambridge for screening by a member of the Cambridge team. Type 4 players will collect their own samples but will then be visited by a researcher from Cambridge to help with pooling and DNA extraction before the samples are sent to Cambridge for genotyping and analysis. GAMES will not inhibit local efforts and will support domestic projects, but players must commit themselves to a spirit of collaboration and, after initial publication, make their results available to the GAMES consortium. Stephen Sawcer reviewed the theoretical background leading to the GAMES experiment and discussed some practical issues. Since this provides the basis upon which GAMES is designed and is therefore of special interest to all participants, his talk is summarised separately and in more detail in Appendix 1. Session 2 Multiple Sclerosis in Europe Chairman: Ian McDonald Annette Oturai represented the Nordic group consisting of seven centres operating in four countries. Activity of the centres is co-ordinated by twice yearly meetings. A central database and DNA bank are located in Copenhagen. The group has received EU grants since the 1980s for the identification of candidate genes/loci . The collection of 200 sib-pairs is almost completed, and a linkage screen is currently 7 ongoing. The collection of 1000 trios has just started, and at least 100 cases and controls have been collected by each centre. Possible susceptibility loci identified in 22 Denmark include two on chromosome 5 (D5S1953, D5S667) and one on chromosome 17 (D17S787). A strong association of multiple sclerosis with HLA DR 2 is confirmed by the Danish group, demonstrating an odds ratio of 3.7. Anne Spurkland reported on the progress of sample collection in Norway, where cases and controls have already been collected for work on HLA23, immunoglobulin regions and T-cell activation genes. A locus of interest is the SHC2D2A gene, which lies close to the CD1 region; its product is expressed on activated T-cells. It is implicated in EAE and some alleles are associated with multiple sclerosis even though there is no linkage to flanking markers. Jan Hillert explained that over 2,300 MS patients have already been collected in Sweden, of which 1,300 are included in a registry. No effect of HLA on disease severity has been demonstrated and a current project compares features of the 96 most benign cases with the 96 most severe ones. Special features of the Finnish population were discussed by Pentti Tienari. Familial clustering and high prevalence rates (i.e. 200/100,000 in Seinajoki) in West Finland make the population particularly interesting for genetic analysis24. A genome screen in multiple sclerosis demonstrated regions of interest on chromosomes 5, 6,17 25 and18 . To date, 1527 families have been identified and 150 trios and 27 multiplex families are collected. Jeff Gulcher represented DeCode, the major centre for genetic studies in Iceland. The Icelandic population offers unique opportunities since it consists of large extended pedigrees for which about 36,000 ancestors are known. Databases of family trees extending back more than five hundred years frequently reveal relationships between individuals, most of whom are unaware of their extended family. A genealogical approach, based on the fact that haplotypes may be identical for these distantly related affected individuals, is currently being employed for the 26 analysis of several complex diseases . Definitions of phenotypes are broad, and phenotype transmission is used to decide which of the extended families will be included in the analysis. Using this approach it has been possible to identify at least one locus with genome-wide significance in stroke patients (ischaemic and haemorrhagic). Psoriasis is another example where a locus with a Lod score of 1.5 before and over 5 after HLA segregation could be identified in 300 related patients. 8 Iceland has currently 330 living Multiple Sclerosis patients, and blood has been collected from 270. A genome-wide scan involving two thirds of the patients has not identified any regions of unequivocal linkage. Marcin Mycko explained that the prevalence of multiple sclerosis in Poland is 90/100,000, with about 14,000 individuals currently affected. The Polish population is genetically homogeneous and there are few minorities. Blood has already been collected from 700 affected individuals, including 50 mutiplex families. About 90 cases have been used for previous candidate gene studies, looking at the HLA region and 2 polymorphisms of ICAM-127. Future studies looking at other candidate genes are underway. The group from Holland was represented by speakers from two centres. Chris Polman from Amsterdam reported that previous attempts to dissect the genetic base of multiple sclerosis have mainly taken the candidate gene approach, looking for cytokine polymorphisms which are associated with clinical course or disability. The 20 interleukin 1 receptor agonist is associated with disease progression ; interleukin 28 10 is associated with TNF production. Gerhard te Meerman from Groningen emphasised that the dissection of complex diseases is compromised by weak effects attributable to the involved genes. He pointed out that accurate determination of allele frequencies is crucial for showing association and that power is limited by features such as allele frequency mismatching. He felt that before 'blowing the start whistle' for GAMES, two or three times as many markers were needed, the artefacts inherent to estimating allele frequencies using pooled DNA had to be resolved, and the possibility of null alleles limiting the usefulness of the markers in different populations checked by direct sequencing of each marker in each population. He proposed examining specific genetic regions in detail rather than performing a whole genome screen. Mapping specific genes in the HLA region, for example, and identifying the disease polymorphism might prove more useful than conducting a potentially under-powered experiment and this should be postponed until appropriate technology is available. 9 Herwig Carton reviewed the epidemiologic features of multiple sclerosis in Belgium 29 . There are 10,000 affected individuals. The relative risk for family members is 12 30 for first degree relatives and 3 for second degree relatives . Sample collection is currently in progress from 675 affected patients and 2400 relatives. Several multiplex families are included in this number. The genetic heterogeneity of multiple sclerosis is reflected by the development of a multiple sclerosis like illness in patients with Leber’s optic atrophy. The LHON registry in Belgium comprises 103 patients of which 5 suffer from Harding’s disease, in which the clinical phenotype is identical to classical multiple sclerosis with the exception that optic nerve involvement is more severe 31. Professor Jörg Epplen talked about work being performed by the Bochum group 32 from Germany. Projects are being conducted on several autoimmune diseases (multiple sclerosis, rheumatoid arthritis, inflammatory bowel disease are vasculitis) in which the same control groups can be used. In Wegener’s Granulomatosis, only 5/10 microsatellite or SNP markers from the proteinase 3 gene region, which is already implicated, show association with the disease. The HLA association in multiple sclerosis illustrates that normal alleles in specific combinations can be 33 important for conveying susceptibility . An association study of the NFB cascade is being performed in over 400 probands. SNPs may show limited polymorphisms. The p50 and p65 gene in Caucasians, for example, has only two SNP polymorphisms associated with the gene. Frauke Zipp from Berlin explained the present lack of interaction between potential GAMES players in Germany. About 100 cases and controls have so far been collected in Berlin, but access to the Serono database may provide an additional 700 cases. Although there is higher TNF, lymphotoxin and IFN production in cases, TNF productions by T cell clones does not correlate with TNF alleles, HLA might influence proinflammatory cytokine production, but so far no correlation has been demonstrated in animal models. 34 Current work in Berlin looks at candidate genes from the apoptosis context , especially those involved in T-cell regulation. Northern Ireland has contributed significantly to the establishment of genetic epidemiologic studies, participating in the UK sib-pair study and urban-rural difference in risk. Stanley Hawkins explained that there is little migration in Northern Island; in the North East, 71% of probands were born in the region and 91% in 10 Northern Ireland. Primary progressive multiple sclerosis accounts for 9– 25% of patients, and has distinct features; less female preponderance, later onset and less MRI lesion-load. There is lack of agreement on what constitutes PPMS; in Ireland this means progression of disability from onset, allowing for episodes of deterioration during inter-current illness. Long-term follow-up studies of patients considered to have benign MS (i.e. EDSS <3 10 years after diagnosis) show that most have progressed significantly 20 years after diagnosis. No HLA association with clinical 35 36 subgroups or disability has been demonstrated . Sample collection in Northern Ireland has already been performed in 174 RR/SP patients, 113 PP patients and 220 controls. Graeme Stewart reviewed the epidemiology of multiple sclerosis in Australia, which has a landmass comparable to the United States excluding Alaska, but only about 20 million inhabitants. The first British settlement was established 1788 and the Federation formed in 1901. Even though the population is multicultural, most inhabitants originate from the United Kingdom or Ireland. About 80% live in inner city areas. Despite a common genetic background, the prevalence of multiple sclerosis shows a striking variation with latitude, ranging from 75/100,000 in the south to 12/100,000 in the sub-tropical areas. No association between risk of developing multiple sclerosis and country of origin has been demonstrated. Blood samples from 200 cases, 200 controls and 206 trio families have been collected; 87% are of Northern European origin and only 2% from outside Europe. An Australian longitudinal database will be established from April 2000 to contain information on about 2000 patients. A workshop planned for June 2000 will further define plans for collecting prospective data and correlating genotypes with clinical aspects. The Tasmanian population, which is ideal for genealogical studies, offers a special opportunity in Australia. Several extended pedigrees have been identified and a genome screen with 300 markers is underway. Michel Clanet spoke for the French group. A national collaboration has existed in France since 1994 with clinicians at 3 centres and a laboratory in Paris. The group is mainly funded by the Multiple Sclerosis Society and so far has collected 261 simplex families and 97 multiplex families. All patients fulfil the Poser or Goodkin criteria and 11 have been personally examined by a member of the group who records EDSS, clinical course, duration and progression. Francoise Clerget-Darpoux debated the most promising strategy for GAMES, acknowledging that power calculations have to assume many unknown factors; the size of the genetic effects sought, the extent of interaction and the presence of heterogeneity are all unsolved; the optimum number of markers is unclear; increasing the number will inflate the number of false positives. Questions about what happens at the end of the "GAME" need to be addressed: who "owns" interesting results when they become available; when and by whom will the meta- analysis be performed; and who plays in what order? Any search for genes is associated with risk, and the GAMES experiment is no exception, but the prospects of success make that risk worthwhile. SESSION 3 Multiple Sclerosis in Southern Europe Chairman : Kees Lucas Luca Massacesi spoke for the group from Northern Italy. A focused linkage based genome screen looking at 14 promising areas in 41 Italian and 28 Sardinian multiplex families was published in 199937. Analysis was separate for the continental and Sardinian data set. Even though, the linkage scores obtained were low, some markers (at 2p11, 3q21.1, 7p15.2 and 22q13.1) stood out as promising. Simon Broadley then performed a linkage-based screen of the whole genome in 40 multiplex families using 322 markers. Regions of interest could be identified on chromosomes 5, 10 and 16, but no area of genome-wide significance was demonstrated. There was no peak in the HLA region, illustrating the low power of linkage studies with relatively small sample sizes. Hits on chromosome 2 and 22 overlap with the UK screen, and others on chromosomes 8 and 10 have been shown to be important in diabetes. A peak on chromosome 7 overlaps with an autoimmunity cluster described by Becker. Luigi Grimaldi described candidate gene studies performed in Milan. Interleukin 1 receptor agonist is associated with occurrence of 12 38 disease and clinical course variability in 339 Italian MS patients . PECAM-1 gene 39 40 and APOE polymorphisms fail to show association with the risk of developing MS or clinical course. Two groups in the Milan area are currently recruiting patients for GAMES. So far blood samples from 325 sporadic cases, 110 trios and 155 multiplex families have been collected. Maria Trojano explained that the prevalence of Multiple Sclerosis Southern Italy is 50/100.000 and currently about 2000 patients, including 300 from multiplex families attend regular clinics. Multiple Sclerosis candidate genes41, already examined, include various cytokines, metalloproteinase and mitochondrial mutations. A corporation of six centres in South Italy aims to register all familial neurological diseases, define predictive laboratory and diagnostic tests and analyse genetic factors. The first phase of the plan includes a network creation where each centre will be connected to an electronic database. Patients will be recruited if they fulfil Poser criteria for definite or probable MS. Patients will be segregated in clinical subtypes during the second phase and clinical events/disease progression recorded 42. The aim is to collect blood from 2000 cases, 1000 parents and 500 controls and perform TDT and case/control studies on 7 microsatellite markers and 200 SNP’s in addition to looking for mitochondrial mutations. Maria Liguori reported that 1200, including 87 multiplex families patients, are currently being followed at the University Hospital of Bari,. For the GAMES experiment, blood has already been collected from 125 MS sporadic cases, 112 parents, 28 sib-pairs and 150 controls. Francesca Coraddu discussed epidemiologic features of Sardinia, where prevalence (150 /100,000) and incidence ( 80/ year) of multiple sclerosis are much higher than in surrounding Mediterranean countries, suggesting that isolated population growth has concentrated susceptibility factors.The special genetic basis of susceptibility is reflected by the established association with DR3 (DRB1*0301, DQA1*0501, DQB1*0301) and DR4 (DRB1*0405, DQA1*0501, DQB1*0302) haplotypes, which are distinct from those seen in Europeans of Caucasian origin 43. 13 A linkage genome screen in 56 multiplex Sardinian families, including 46 sibling pairs, 3 sibling trios and 7 parent child pairs, screened with 327 microsatellite markers,demonstrated no region of genome-wide significance. However, regions of interest were identified on Chr 1q,Chr 10q and Chr 11p where the MLS reaches 1.8. The "Regional Multiple Sclerosis Centre" in Cagliari has established a DNA bank for Southern Sardinia. At present 551 trios families, 62 multiplex families and 190 sporadic cases have been collected . Giovannie Savettieri represented Sicily, where epidemiological surveys have been conducted in 5 centres including follow-up studies in 3. Enna was found to have prevalence rates of 53/100 000 in 1975 compared to 113.3/100.000 in 1995; 44 45 prevalence was 43/100 000 in 1981 and 72.4/100 000 in 1998 in Monreale , 32/100.000 in 1975 and 51/100 000 in 198146 in Agrigento and 50/100.00 in Caltanisetta City and Baghara in 1980 and 1994 respectively. No genetic studies have yet been been performed in Sicily. The gradient in prevalence which contrasts with Malta (4/100.000) is unexplained. Xavier Montalban from Barcelona indicated that prevalence rates of 53-65/100 000 have been established Spain in the last 10 years. A collaboration of 3 neurological centres which includes 25 Neurologists and 3 DNA Labs will be coordinated from Barcelona (UNIC). The group in Barcelona, Madrid and Valencia. Funding is available from the Spanish Multiple Sclerosis Society. Pablo Villoslada reported that blood for GAMES has already been collected from 265 Simplex families, 25 Multiplex families, 200 cases and 200 controls. Special features of Spain which may be interesting for genetic research are the Canary and Baleares Islands and several partially genetic isolated areas( the Basque area in the North and Galicia in the Northwest). . 14 Berta Silva reviewed the epidemiology of multiple sclerosis in Portugal. There is a national register of patients aiming to study the epidemiology of Multiple Sclerosis. Prevalence was found to be 30/100.000 in 1996 and 49/100.000 in 1998. In preparation for GAMES, blood samples from about 100 cases have already been collected. Further collection of cases, controls and trio families is planned and will be facilitated by recruiting probands from the National Register. . Pimpi Reyes-Yanez discussed epidemiologic features of the Canary Islands. The population is Caucasian with genetic influences from Spaniards, Romans, Vikings, Africans and others. The prevalence of Multiple Sclerosis was 6.8/100 000 in Las Palmas in 1983 and of 18.3/100 000 in Lanzarote in 1987. The difference is probably methodological but both regions show lower prevalence compared to mainland Spain 47 (50/100 000). Susceptibility is associated with DR15 and also DR4 . The DR4 haplotype is different from unique Sardinian DR4 , but the same as the HLA DR4 associations seen in Jordan, Malaga or Turkey. This association may represent genetic input from the Eastern Mediterranean. 48 John Milonas explained that studies in Northern Greece between 1970 and 1984 demonstrated an incidence of 1.8/100 000/yr and prevalence of 29/100 000. Further studies are needed, because the figures are probably higher, reflecting better knowledge of the disease and better means of investigation. About 680 patients with clinically definite MS (9% of which have a benign course) characterised in Thessaloniki between 1990 and 2000 could potentially be recruited for GAMES. Collaborations exist with centres in Crete, Athens and Padras who are also recruiting patients, and blood collection is due to start next month. Mefkure Eraksoy reported that the earliest publications on Multiple Sclerosis in Turkey date from 1934 when Uzman suggested that the disease only occurs in 15 those who visit Europe. Further studies in 1949, 1954 and 1998 established that clincal presentation and prognostic factors are similar to those seen in other 49 European countries . The Turkish MS Study Group was established in 1993. Currently 1075 patients with clinically definite, laboratory supported clinically definite or laboratory supported probable MS are being followed in Istanbul. Patients are seen mainly in two large hospitals, but also by 249 neurologists practising in the city. At the University Hospital in Istanbul, 5% of neurological patients have MS, and about 20-25% are newly diagnosed per year. Female/Male ratio is 2:1 and 3.2% are multiplex families. Special epidemiological features of Turkey includes the high rate of consanguineous marriages (1:5). The first epidemiology studies of MS in Cyprus were undertaken in the early 1990’s. Koulis Kyriallis reported that similar prevalence figures were found in the Greek 50 and Turkish communities . Middleton and Dean found prevalence rates of 45/100 000 (similar to Sicily) in 3 rural parts of Cyprus in 1991. More recent numbers from 1999 show a prevalence 67/100 000 for clinically definite MS in the Greek population of the island, with 15% constituting familial cases. The increased incidence and prevalence figures in the last 30 years are probably due partly to increased use of MRI imaging technology. However, preliminary data suggest that incidence continues to rise so that prevalence rates may come close to 100/100 000, as in northern European countries. Demographic features of MS patients seem to be similar to other European countries. Blood has already been taken and DNA extracted from 296 patients, 290 parents and 200 controls. Multiple sclerosis seems to be rare in Arabs and the earliest reports date back to 1958 when studies were carried out in Iraq. Amir Al-Din explained that more recent 51 studies studies in Kuwait suggest that prevalence rates for Palaestinians (23/ 100.000) may be higher compared to other Arabs (13/100.000), particularly those from the Gulf (7.5/100.000). These figures are not influenced by migration of Palaestinians during the wars of 1948 and 1957, and are similar in Jordan (prevalence of Palaestinians 33/100.000 compared to 16/100.000 in Jordanians)52. Studies on HLA susceptibility factors in Palaestinians53 show association with DR2, DR3, and DQw1. DR1 and DR5 may have protective effects, as has already been suggested in Kuwait where all patients were negative for DR1. The lack of a DR2 16 association in Kuwaiti Arabs is probably due to high prevalence of this allele in controls. Multiple sclerosis in Jordan has different demographic features compared to European countries. The country has a much younger population with a 60% increase in < 14 year olds, creating a pyramidal age-stratification curve. There are no other differences regarding prognosis, imaging and phenotype. The King Hussein Medical Centre is the main neurological centre in Amman (Jordan) with a register of 800 patients. Information on some pedigrees still needs documentation but blood collection, DNA extraction and pooling could be done locally. SESSION 4 A consensus on how to play GAMES Chairman : Professor I. McDonald Mefkure Eraksoy talked about her experiences whilst organising the Turkish GAMES group. She visited Cambridge for several weeks, to obtain more information about GAMES and find out how the work is organised. Whilst still in Cambridge, she formed of the Turkish Multiple Sclerosis Study Group (TMSGSG) which involves 34 neurological centres, at least one of which is located in each region of Turkey. The majority involves University Hospitals. Funding for the experiment and ethical approval are currently being applied for. Blood has already been collected from 200 controls, 180 cases and 180 trio families. Information about specific family structures is being recorded and this is especially important in view of the high rate of 17 consanguinous marriage in Turkey. Professor Eraksoy said, that she encountered no specific problems when recruiting centres for the experiment. Mara Giordano discussed practical issues of pooling DNA samples. As a proof of principle, she described an experiment pooling DNA from 400 individuals. DNA was quantitated by Pico Green Fluorescent Assay and diluted to a final concentration of 20ng/ul. The pools contained 200 ul (4ng) of each DNA and 5 ul (100 ng) of the pool was used for each PCR. Pools of different sizes were created after dividing the 400 individuals into 5 subgroups. In the first stage experiment, each subgroup pooled together was analysed; in the second stage, 4 pools each containing 2 subgroups; in the third stage 3 pools each containing 3 subgroups; and finally one pool of all the individuals together. When comparing results with those achieved by individual genotyping, it could be demonstrated that reliability increased with pool size, due to a reduction of experimental and sampling errors. In the discussion that followed, Professor Lucas suggested that the Cambridge group should provide a detailed protocol describing methods to be used for pooling DNA. Stephen Sawcer pointed out that it would not be possible (or appropriate) to be prescriptive since facilities available to each group vary. He pointed out that several methods for constructing pools have been used and are published (that suggested in appendix 4 is based on the existing literature).. Each group should test their own pool using a positive control such as the TNFa microsatellite which is known to be in LD with HLA or an SNP, previously typed individually, should be typed in the samples pool. Although checking the validity of pools is appealing, it is not clear how validity checking might be performed and the tolerances which would be considered appropriate are undefined, as are actions to take in the event of failure to meet these tolerances. Clinical classification, definition of cases and the possible inclusion of subgroups is another crucial issue of practical importance for GAMES. Michel Clanet stressed the importance of strict inclusion criteria, especially in view of the fact that up to 10% rate of mis-diagnosis in families. Cases classified as definite multiple sclerosis should fulfil the following 6 criteria: - at least one objective abnormality on clinical examination, - predominant involvement of at least one CNS long tract - two or more episodes or stepwise progression with abnormal CSF - two or more CNS areas invoved ( including paraclinical evidence) 18 - age between 10 and 60 years - alternative diagnosis excluded by appropriate tests Only probands with definite or probable MS (5 out of the 6 criteria) should be included. Most groups have used the Poser54 or the Goodkin55 criteria for previous studies. The need for longitudinal evaluation of cases, i.e. use of progression indices measuring EDSS per disease duration, should be addressed. The question of stratification on clinical course arises, particularly when separating primary progressive cases from other forms of the disease. Chris Polman drew attention to the fact that any sub grouping invariably leads to loss of power. However if the primary progressive group is to be separated, strict criteria must be employed. In the ensuing discussion, it became clear that most GAMES players would be in favour of separating primary progressive cases. Ian McDonald mentioned that new criteria for the diagnosis of primary PPMS are due to be published in the Annals of Neurology this year. Inclusion in the GAMES PPMS subgroup would require patients to be definite PPMS (includes abnormal CSF) or probable PPMS (may allow for normal or not available CSF result) according to these criteria. Benign cases should be included in the relapsing remitting/secondary progressive group. Oluf Andersen addressed ethical considerations. There is general agreement about informed consent and ethics local committee approval as the basic requirement for any study involving patients. The exact guidelines to be followed vary between local authorities. The Swedish Research Council requires consent for storage of DNA and informed consent for each new purpose (a general acceptance for unspecified causes is not acceptable). Exceptions are studies which closely resemble the original aim. Information must be given on methods of storage and storage time, and donors must be informed that storage can be terminated at their request. International collaborations require that only coded material is distributed. Jeff Gulcher emphazised the special ethical considerations in a small country where researchers may know participating patients56. To ensure collaboration by hospitals and general practitioners, a strict system of data protection is required. In Iceland, a third party system exists where neither the participating doctors nor researchers have access to the key necessary to decode patient numbers. Researchers requiring clinical information must contact the third party who makes the information available to them57. Bar codes rather than numbers on blood samples provide an additional level 19 of security. In the course of discussion, it was decided that the third party system would be good for the GAMES experiment. Ian McDonald agreed that the International Federation of Multiple Sclerosis Societies might be a suitable institution and offered to discuss this internally. Alastair Compston initiated the final discussion of the first GAMES meeting by reviewing several issues that had been raised. Whilst the experiment may lack power, there are a lot of unknown factors necessary to make this judgement and GAMES is the most powerful experiment that can currently be performed. On the question of using either SNP based genome screens or a dense microsatellite map, the latter is currently available and GAMES already uses 85 - 90% of informative microsatellites. GAMES provides a unique opportunitiy to look at areas of the genome which have never been systematically examined in several countries. Taking a straw poll of all those present at the meeting it appeared that the vast majority wanted to participate. The approximate number of type 1, type 2, type 3 and type 4 players was 8,12,6 and 2 respectively. One of the practicalities for GAMES is resources. Substantial funding for the 6000 microsatellite markers, the technical equipment and the screening of the UK population has already been provided in Cambridge by the Wellcome Trust. Application for an additional set of 400 markers for linkage studies in the Scandinavian, Turkish and Australian population has been made to the MS Societies of Great Britain and Northern Ireland. Application to the European Community for funding a Marie Curie host institution are currently ongoing. If successful this would provide a permanent salary for type 2 players to cover their stay in Cambridge. If a type 2 player comes with funding, the Mary Curie salary could be distributed to other GAMES players. Costs not covered by this scheme would need to be applied for locally by the participating groups, such as those needed for sample collection, DNA extraction and pooling. Type 1 players would need to cover the running costs of their lab and genotyping, although markers will be provided by Cambridge, at the cost of plates onto which they are aliquoted. The issue of how individual groups should best apply for funding was raised. The option of applying individually carries the disadvantage that several groups might compete for the same grants. The alternative is to apply as a consortium. Michel Clanet was in favour of the latter and offered to co-ordinate such an initiative. 20 International collaborations are prone to cause conflicts of interest and it is important that the conditions of GAMES are transparent to all players. All groups are free to publish their own results and follow up interesting findings resulting from their screen. Results obtained by each group will be made available to the rest of the team after a suitable interval and included in the meta - analysis. Evolving data will be published on a GAMES website. It still needs to be decided if all players or only those who have made already made their data available would have access. It is not required that other GAMES players are co-authors in interim publications (although the Wellcome trust must be acknowledged). The meta-analysis will be published by the GAMES group, not individual authors. A schedule for type 2 players will have to be worked out and some groups may want to change to type 1 or type 3 in order to move faster. One rate limiting step for the experiment is access for type 2 players wishing to visit Cambridge. Possible solutions include recruiting a second centre. Iceland offered this facility. A second alternative is to purchase a second ABI 3700. . Stephen Sawcer pointed out some practical issues raised in the meeting, including the theoretical possibility of multiplexing, the arrangement of plates to markers with similar PCR conditions, and the allocation of chromosome X markers on separate blocks. Ian McDonald closed the first GAMES meeting by wishing everybody good luck and demonstrating true sportsmanship by being the first GAMES player to be seen wearing the official T-shirt. REFERENCES. 1 Sawcer S, Jones HB, Feakes R, Gray J, Smaldon N, Chataway J, Robertson N, Clayton D, Goodfellow PN, Compston A. A genome screen in multiple sclerosis reveals susceptibility loci on chromosome 6p21 and 17q22.Nat Genet. 1996 Aug;13(4):464-8. 2 Ebers GC, Kukay K, Bulman DE, Sadovnick AD, Rice G, Anderson C, Armstrong H, Cousin K, Bell RB, Hader W, Paty DW, Hashimoto S, Oger J, Duquette P, Warren S, Gray T, O'Connor P, Nath A, Auty A, Metz L, Francis G, Paulseth JE, Murray TJ, Pryse-Phillips W, Risch N, et al. A full genome search in multiple sclerosis.Nat Genet. 1996 Aug;13(4):472-6. 3 Haines JL, Ter-Minassian M, Bazyk A, Gusella JF, Kim DJ, Terwedow H, Pericak-Vance MA, Rimmler JB, Haynes CS, Roses AD, Lee A, Shaner B, Menold M, Seboun E, Fitoussi RP, Gartioux C, Reyes C, Ribierre F, Gyapay G, Weissenbach J, Hauser SL, Goodkin DE, Lincoln R, Usuku K, Oksenberg JR, et al. A complete genomic screen for multiple sclerosis underscores a role for the major histocompatability complex. The Multiple Sclerosis Genetics Group.Nat Genet. 1996 Aug;13(4):469-71. 21 4 Coraddu F, Sawcer S, Feakes R, Chataway J, Broadley S, Jones HB, Clayton D, Gray J, Smith S, Taylor C, Goodfellow PN, Compston A. HLA typing in the United Kingdom multiple sclerosis genome screen.Neurogenetics. 1998 Dec;2(1):24-33. 5 Mein CA, Barratt BJ, Dunn MG et al. Evaluation of Single Nucleotide Polymorphism typing with invader onPCR amplicons and its automation. Genome Res 2000; 10(3): 330 - 343 6 Germer S, Holland MJ, Higuchi R.High-throughput SNP allele-frequency determination in pooled DNA samples by kinetic PCR.Genome Res. 2000 Feb;10(2):258-66. 7 Risch N , Merikangas K. The future of genetic studies of complex human diseases. Science 1996; 273: 15156- 1518 8 Okamoto T, Suzuki T, Yamamoto N. Microarray fabrication with covalent attachment of DNA using Bubble Jet technology.Nat Biotechnol. 2000 Apr;18(4):438-441. 9 Brenner S, Williams SR, Vermaas EH, Storck T, Moon K, McCollum C, Mao JI, Luo S, Kirchner JJ, Eletr S, DuBridge RB, Burcham T, Albrecht G.In vitro cloning of complex mixtures of DNA on microbeads: physical separation of differentially expressed cDNAs.Proc Natl Acad Sci U S A. 2000 Feb 15;97(4):1665-70. 10 Risch N, Teng J. The relative power of family-based and case-control designs for linkage disequilibrium studies of complex human diseases I. DNA pooling.Genome Res. 1998 Dec;8(12):1273-88. Review. 11 Clayton D, Jones H. Transmission/Disequilibrium Tests for Extended Marker Haplotypes. Am J Hum Genet. 1999 Oct;65(4):1161-1169. 12 Jersild C, Svejgaard A, Fog T, Ammitzboll T. HL-A antigens and diseases. I. Multiple sclerosis. Tissue Antigens. 1973;3(4):243-50. 13 Spurkland A, Celius EG, Knutsen I, Beiske A, Thorsby E, Vartdal F.The HLA-DQ(alpha 1*0102, beta 1*0602) heterodimer may confer susceptibility to multiple sclerosis in the absence of the HLA-DR(alpha 1*01, beta 1*1501) heterodimer.Tissue Antigens. 1997 Jul;50(1):15-22. 14 Madsen LS, Andersson EC, Jansson L, krogsgaard M, Andersen CB, Engberg J, Strominger JL, Svejgaard A, Hjorth JP, Holmdahl R, Wucherpfennig KW, Fugger L.A humanized model for multiple sclerosis using HLA-DR2 and a human T-cell receptor.Nat Genet. 1999 Nov;23(3):343-7. 15 Lin L, Faraco J, Li R, Kadotani H, Rogers W, Lin X, Qiu X, de Jong PJ, Nishino S, Mignot E. The sleep disorder canine narcolepsy is caused by a mutation in the hypocretin (orexin) receptor 2 gene. Cell. 1999 Aug 6;98(3):365-76. 16 Becker KG, Simon RM, Bailey-Wilson JE, Freidlin B, Biddison WE, McFarland HF, Trent JM. Clustering of non-major histocompatibility complex susceptibility candidate loci in human autoimmune diseases. Proc Natl Acad Sci U S A. 1998 Aug 18;95(17):9979-84. 17 Ligers A, Xu C, Saarinen S, Hillert J, Olerup O.The CTLA-4 gene is associated with multiple sclerosis.J Neuroimmunol. 1999 Jun 1;97(1-2):182-90 18 Harbo HF, Celius EG, Vartdal F, Spurkland A.CTLA4 promoter and exon 1 dimorphisms in multiple sclerosis.Tissue Antigens. 1999 Jan;53(1):106-10. 19 Evangelou N, Jackson M, Beeson D, Palace J.Association of the APOE epsilon4 allele with disease activity in multiple sclerosis.J Neurol Neurosurg Psychiatry. 1999 Aug;67(2):203-5 20 Schrijver HM, Crusius JB, Uitdehaag BM, Garcia Gonzalez MA, Kostense PJ, Polman CH, Pena AS. Association of interleukin-1beta and interleukin-1 receptor antagonist genes with disease severity in MS.Neurology. 1999 Feb;52(3):595-9. 21 Mertens C, Brassat D, Reboul J et al. A systematic studt of oligodendrocyte growth factor candidats for genetic susceptibility to MS. French Multiple Sclerosis Study Grouip. Neurology 1998; 51 (3): 748 - 53 22 Oturai A, Larsen F, Ryder LP, Madsen HO, Hillert J, Fredrikson S, Sandberg-Wollheim M, Laaksonen M, Koch-Henriksen N, Sawcer S, Fugger L, Sorensen PS, Svejgaard A. Linkage and association analysis of susceptibility regions on chromosomes 5 and 6 in 106 Scandinavian sibling pair families with multiple sclerosis.Ann Neurol. 1999 Oct;46(4):612-6. 23 Spurkland A, Celius EG, Knutsen I, Beiske A, Thorsby E, Vartdal F. The HLA-DQ(alpha 1*0102, beta 1*0602) heterodimer may confer susceptibility to multiple sclerosis in the absence of the HLA-DR(alpha 1*01, beta 1*1501) heterodimer.Tissue Antigens. 1997 Jul;50(1):15-22. 22 24 Sumelahti M, Tienari PJ, Wikstrom J, Palo J, Hakama M,Regional and temporal variation in the incidence of multiple sclerosis in finland 1979-1993.Neuroepidemiology. 2000 Mar- Apr;19(2):67-75. 25 Kuokkanen S, Gschwend M, Rioux JD, Daly MJ, Terwilliger JD, Tienari PJ, Wikstrom J, Palo J, Stein LD, Hudson TJ, Lander ES, Peltonen L. Genomewide scan of multiple sclerosis in Finnish multiplex families.Am J Hum Genet. 1997 Dec;61(6):1379-87. 26 Kong A, Gulcher J, Stefansson K. Genealogy certainly matters for multifactorial genetic disease.BMJ. 1999 Aug 28;319(7209):578-9. 27 Mycko MP, Kwinkowski M, Tronczynska E, Szymanska B, Selmaj KW.Multiple sclerosis: the increased frequency of the ICAM-1 exon 6 gene point mutation genetic type K469. Ann Neurol. 1998 Jul;44(1):70-5. 28 29 Carton H. Multiple sclerosis. A historical review with emphasis on the last 20 years. Acta Neurol Belg. 1996 Sep;96(3):224-7. No abstract available. 30 Carton H, Vlietinck R, Debruyne J, De Keyser J, D'Hooghe MB, Loos R, Medaer R, Truyen L, Yee IM, Sadovnick AD. Risks of multiple sclerosis in relatives of patients in Flanders, Belgium. J Neurol Neurosurg Psychiatry. 1997 Apr;62(4):329-33. 31 Harding AE, Sweeney MG, Miller DH, Mumford CJ, Kellar-Wood H, Menard D, McDonald WI, Compston DA. Occurrence of a multiple sclerosis-like illness in women who have a Leber's hereditary optic neuropathy mitochondrial DNA mutation. Brain. 1992 Aug;115 ( Pt 4):979-89. 32 Epplen C. [Genetic components in autoimmune diseases]. Internist (Berl). 1999 May;40(5):469-75. Review. German. 33 Jaeckel S, Epplen JT, Kauth M, Miterski B, Tschentscher F, Epplen C. Polymerase chain reaction-single strand conformation polymorphism or how to detect reliably and efficiently each sequence variation in many samples and many genes. Electrophoresis. 1998 Dec;19(18):3055-61. Review. 34 Zipp F, Beyer M, Gelderblom H, Wernet D, Zschenderlein R, Weller M.No induction of apoptosis by IFN-beta in human antigen-specific T cells.Neurology. 2000 Jan 25;54(2):485-7. 35 McDonnell GV, Hawkins SA. Major histocompatibility complex class II alleles and the course and outcome of MS.Neurology. 1999 Sep 11;53(4):893-4. 36 McDonnell GV, Mawhinney H, Graham CA, Hawkins SA, Middleton D.A study of the HLA- DR region in clinical subgroups of multiple sclerosis and its influence on prognosis.J Neurol Sci. 1999 May 1;165(1):77-83. 37 D’Alfonso S, Nistico L, ZavattariP et al. Linkage analysis of multiple sclerosis with vcandidate markers in Sardinian and Italian families. Eur J Hum Genet 1999; 7 (3): 377 - 385 38 Sciacca FL, Ferri C, Vandenbroeck K. Relevance of interleukin 1 receptor agonist intron 2 polymorpohisms in Italian MS patients. Neurology 1999; 52; 1096 - 1898 39 Sciacca FL, Ferri C,, D’Alllfonso S et al. Association study of a new polymorphism in the PECAM-1 genome with multiple sclerosis. J Neuroimmunol 2000; 104(2); 174 - 178 40 Ferri C, Sciacca FL, Veglia F et al. APOE epsilon 2-4 and –491 polymorphisms are not associated with MS. Neurology 1999; 53(4): 888 – 889 Session 1 41 Trojano M, Liguori M, De Robertis F, Stella A, Guanti G, Avolio C, Livrea P. Comparison of clinical and demographic features between affected pairs of Italian multiple sclerosis multiplex families; relation to tumour necrosis factor genomic polymorphisms.J Neurol Sci. 1999 Jan 15;162(2):194-200 42 Trojano M, Avolio C, Manzari C, Calo A, De Robertis F, Serio G, Livrea P. Multivariate analysis of predictive factors of multiple sclerosis course with a validated method to assess clinical events.J Neurol Neurosurg Psychiatry. 1995 Mar;58(3):300-6. 43 Marrosu MG, Murru MR, Costa C et al. Multiple Sclerosis in Sardinia is associated and in linkage disequilibrium with HLA DR3 and DR4 alleles. Am J Hum Genet 1997; 61(2): 454 -7 44 Savettieri G, Daricello B, Giordano D, Karhausen L, Dean G. The prevalence of multiple sclerosis in Sicily. I: Monreale city. J Epidemiol Community Health. 1981 Jun;35(2):114-7. 45 Savettieri G, Salemi G, Ragonese P, Aridon P, Scola G, Randisi G. Prevalence and incidence of multiple sclerosis in the city of Monreale, Italy.J Neurol. 1998 Jan;245(1):40-3. 23 46 Dean G, Savettieri G, Giordano D, Butera C, Taibi G, Morreale S, Karhausen L. The prevalence of multiple sclerosis in Sicily. II: Agrigento city. J Epidemiol Community Health. 1981 Jun;35(2):118-22. 47 Coraddu F, Reyes-Yanez MP, Parra A, Gray J, Smith SI, Taylor CJ, Compston DA. HLA associations with multiple sclerosis in the Canary Islands. J Neuroimmunol. 1998 Jul 1;87(1-2):130-5. 48 Milonas I, Tsounis S, Logothetis I. Epidemiology of multiple sclerosis in northern Greece. Acta Neurol Scand. 1990 Jan;81(1):43-7. 49 Kantarci O, Siva A, Eraksoy M, Karabudak R, Sutlas N, Agaoglu J, Turan F, Ozmenoglu M, Togrul E, Demirkiran M. Survival and predictors of disability in Turkish MS patients. Turkish Multiple Sclerosis Study Group (TUMSSG).Neurology. 1998 Sep;51(3):765-72. 50 Dean G, Aksoy H, Akalin T, Middleton L, Kyriallis K. Multiple sclerosis in the Turkish- and Greek- speaking communities of Cyprus. A United Nations (UNHCR) Bicommunal Project. J Neurol Sci. 1997 Feb 12;145(2):163-8. 51 al-Din AS, Khogali M, Poser CM, al-Nassar KE, Shakir R, Hussain J, Behbahani K, Chadha G. Epidemiology of multiple sclerosis in Arabs in Kuwait: a comparative study between Kuwaitis and Palestinians.J Neurol Sci. 1990 Dec;100(1-2):137-41. 52 al-Din AS, el-Khateeb M, Kurdi A, Mubaidin A, Wriekat A, al-Shehab A, Khalil RW. Multiple sclerosis in Arabs in Jordan.J Neurol Sci. 1995 Aug;131(2):144-9. 53 Al-Din AS, Al Saffar M, Siboo R, Behbetani K. Association between HLA-D region epitopes and multiple sclerosis in Arabs. Tissue Antigens. 1986 Apr;27(4):196-200.. 54 Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, Johnson KP, Sibley WA, Silberberg DH, Tourtellotte WW. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol. 1983 Mar;13(3):227-31. 55 Goodkin DE, Doolittle TH, Hauser SS, Ransohoff RM, Roses AD, Rudick RA. Diagnostic criteria for multiple sclerosis research involving multiply affected families. Arch Neurol. 1991 Aug;48(8):805-7 56 Gulcher JR, Stefansson K.Ethics of population genomics research.. Nature. 1999 Jul 22;400(6742):307-8. 57 Gulcher J, Stefansson K. An Icelandic saga on a centralized healthcare database and democratic decision making.Nat Biotechnol. 1999 Jul;17(7):620.