"Multiple Sclerosis and Gender”

Background information, report of lectures

Let me give you a summarized overview of the gender topic in MS and why gender has become such an important issue in MS. Gender is very important in various topics of MS, such as epidemiology, clinical evolution, environmental factors, experimental mechanisms, pathology and treatment. So clearly, the issue of gender covers a full spectrum of MS.

The Canadian Collaborative Project on the Genetic Susceptibility of MS (CCPGSMS) started in 1993 is now entering its 5^th phase. It has been funded by the MS Society of Canada Scientific Research Foundation. Its database contains more than 29,000 unique families of which at least one member has MS. Also there are more than 8,000 ‘spouse' controls and over 4,000 families with at least two members with MS. This is a very unique database with a number of strengths. These include the longitudinal nature (e.g. maternal effect, prevalence changes), the ability to have controls (spouse controls and sib controls) and the scope on ethnic diversity. These have been very important in helping us to identify many of the gender issues. It is a quite unique study, which has not been duplicated anywhere in the world. This has to do with the fact that Canada has a unique MS clinics network that has a long history of working together.

Historical background

Females get MS more often than males. The history of gender ratio shows that this has not always been the case.

Between 1870 and 1910 it was thought that males may have had MS more often than females or perhaps that men and women had MS at the same rate. These observations were mainly based upon anecdotal evidence from neurologists experience. In the 1920's it was reported by Weschler (n=1,505) that the male sex is more subject to MS than the female in the proportion of approximately 3 to 2. In the 1930's and 1940's it was no longer believed that MS was more common in males. There was an official statement by the American National MS Society declaring equal sex ratio between men and women in MS. It was in the 50's and 60's that we started to think again about a difference in gender. It was then thought that the disease occurred perhaps more often in females. In the 1970's it was finally determined that for every male there were 1.4 females who might get MS. More recently (80's and 90's) it has been found that the sex ratio is in fact 2:1, meaning that for every two females there is only 1 male who gets MS. However, this ratio may vary in different geographic regions. Nowadays, the gender ratio seems to be changing very dramatically. There's speculation that the gender ratio is on the rise. Also gender ratio has proved an important tool to establish whether the actual MS rate is changing. It is often thought that the overall rate of MS is increasing. But it has always been difficult to establish whether this is a true increase or whether this is an observation affected by variables such as earlier age of onset, diagnosis at earlier stages of the disease or better imaging techniques (such as MRI). One way of looking at this is to study the sex ratio's by years of birth. Any artefacts of time (such as advances in diagnosis) should affect males and females equally. These days the gender ratio is approaching 3.2 females to 1 male. Within such a short period of time the rapid risk increase among females cannot be attributed to changes in genetic factors. Therefore, there has to be something in the female environment that is changing and increasing the risk in females but not in males.

Age of onset

Another way of looking at gender is the onset of MS and differences in the gender ratio depending on when you get MS and who gets it. Available data on MS onset in childhood have showed a female preponderance among paediatric MS cases varying between 1.3 and 3.0, largely depending on the age of onset when MS was diagnosed. Also a female preponderance has been found in the conversion from paediatric acquired demyelination syndromes (ADS) to MS. This means that the female gender gives a higher chance of evolving from ADS to MS.

The peak age of onset in MS is 24 years in females and 25 years in males. Clear differences in age of onset were found between sporadic MS cases (no prior MS cases within family) and familial MS cases (earlier reports of MS within family). Overall, the mean age of onset was older among sporadic cases compared to familial cases, though in absolute terms the difference is less than one year. Sporadic males had a significantly older mean age of onset compared to familial males (difference only 1.5 years). Sporadic females showed a trend toward an older mean age of onset compared to familial females (about 6 months, not significant).

When looking at the age of onset of MS among family members the following preliminary observations (not yet published) were made. If a family member who is a twin or a maternal ½ sibling develops MS, their family members who are the co-twin or the other ½ sibling have a risk of developing MS much closer to the age of onset of the first person. So for example, if one member of an identical twin got MS at age 22 and the co-twin doesn't get MS within 7 years time, then this co-twin is not getting MS. This is important information in risk counselling. The highest correlation to age of onset was found in twin pairs (both DZ and MZ pairs). It is important to consider that these twin pairs have been in the same intrauterine environment at the same time, suggesting that there may be a maternal effect on the age of onset. Also a closer correlation to age of onset was found among maternal ½ siblings than paternal ½ siblings. While having the same amount of genetic material, these siblings differ in the way that maternal ½ siblings have a shared maternal environment whilst the paternal ½ siblings have not. These data together with other information are coherent with the observation that there seems to be a maternal effect, possibly an intrauterine environmental effect on the risk of developing MS. So, the gender of the parent may be an important issue.

Gender and recurrence risks

Sisters of female MS patients have a much higher risk of developing MS than brothers. This suggests a correlation between recurrence risk and same gender when the first MS patient is a female. However, if the first member of a family to develop MS is a male, no difference in recurrence risks between genders was observed. This suggests a role of gender in recurrence risk. The importance of understanding gender effect lies not only in risk counselling, but also in developing preventive therapies.

Recently, offspring of MS affected fathers were reported to be more likely to have MS than those of affected mothers. This was attributed to the Carter effect which is seen in polygenetic disorders. It predicts that an affected parent of the sex lesser affected by the disease or trait is more genetically loaded for risk alleles and thus transmits this more often to their offspring. This hypothesis was tested in a population based Canadian MS cohort. The results showed an equal transmission of MS from affected fathers compared to affected mothers (9.41% vs. 9.76%) even after stratifying by sex of affected offspring. Therefore, the conclusion was that there is no evidence for the Carter effect of polygenic heritance on MS susceptibility. This implies that the gender of the MS parent doesn't influence the risk of developing MS in the offspring.

Gender and the X chromosome

From the Canadian cohort a study was performed among 552 sibling pairs and 195 aunt-uncle / niece-nephew (AUNN) pairs. Parent-of-origin effects were explored by dividing AUNN pairs into maternal and paternal. The data showed very clearly that there seems not to be a relationship with the X chromosome. It was concluded from the study that it is unlikely that the X chromosome harbours an independent susceptibility locus or one which interacts with the HLA. This rules the X chromosome out as a factor in the gender effect found in MS.

Gender corrections and MS risk factors

When looking at risk factors for MS and gender effect the following can be said. Only when stratifying by sex, it becomes clear that birth weight and gestational age are no risk factors for MS. For both birth weight and gestational age are influenced by gender. Also no significant differences were found between MS cases and controls for all viral exposures (Infectious Mononucleosis excluded) and vaccinations after stratifying for sex. These examples show that gender stratification is important because there are more females affected with MS

than males. So when you are using controls, you will tend to have more males in the control group (especially in spouse control) and thus bias your results by gender.

The sex-specificity of recall and reporting bias is another type of gender related ascertainment bias. The greater female awareness of medical history has long been qualitatively known to experienced practitioners. Two studies from the Canadian cohort have shown that males, regardless of whether they were index cases of spousal controls, were consistently more unsure of their infection or vaccination history than females. Also males significantly under-report disease in themselves and their first degree relatives as compared to females. Taken together, the data from the CCPGSMS are the first to both validate and quantify this in MS.

MS, gender and reproduction

With regard to reproduction there has been a lot of interest in females with MS but virtually no interest in males with MS. Even though females get MS three times more often than males, there is still a large population of males with MS. So even a topic of reproduction has to be important for both genders. There have been studies that have focused mainly on females and reproduction. These looked at relapse rates and the long term course of MS in females who had MS before pregnancy versus females who didn't have pregnancies during MS. But still there is a lack of information despite very good attempts of documenting reproduction in MS. There is very little information for both genders on SPMS or PPMS and pregnancy. There is little correlation of relapse rates and EDSS prior to and at conception compared to pregnancy and post-partum. There is also the issue of pseudo vs. true relapses, fertility vs. choices, what to do with disease modifying therapies and washout periods and resumption (during gestation, breast feeding), and what influences reproductive decisions.

When considering pregnancy and MS related issues there are large gender differences. Females are largely concerned about the way that MS is going to affect their parenting, they are concerned about being a burden to their partner and about medications and therapies. Males are more concerned about the risk for their child of developing MS and problems with sexual functioning. They are far less concerned about fatherhood parenting or being a burden to their partner. Males do not seem to be as aware as females about pharmaceutical issues that may affect sperm.

Decisions to have children after diagnosis of MS do not differ significantly between males and females with MS. 54.0% of females have one child after MS is diagnosed, versus 54.5% of males. 36.2% of females have two children after MS diagnosis versus 35.1% of males, and 7.7% of females have 3 children versus 7.1% of males.

We are working together with the Canadian MS Society on building a website on reproductive counselling for both males and females diagnosed with MS. This is to assist MS patients in their decision to have children and the issues they have to think about in their decision making process. The topics being covered are the effects of MS on pregnancy (including conception, sexual functioning, pregnancy management, delivery and pregnancy outcome), effect of pregnancy on MS (short- and long-term effects), risk to have a child who will develop MS, teratogenicity of MS treatments, pregnancy outcome and psychosocial issues.

Epidemiology

CANDORE, Palermo, Italy: Gender related immuno-inflammatory factors and human diseases

Sex hormones and immuno-inflammatory response

Estrogens have anti-inflammatory and pro-inflammatory roles depending on influencing factors. Estrogens influence immune response via important pro- and anti-inflammatory pathways (via cytokines) in different cell types, such as B-cells, T-cells, dendritic cells, monocytes/macrophages, macro- and microglia and fibroblasts. Important pro-inflammatory cytokines (such as IL-6, IL-8 and TNF) are typically inhibited at high E2 levels as seen during peri-ovulation and pregnancy. Low E2 concentrations have demonstrated no effect or even stimulatory effects. This renders a woman in the postmenopausal phase more vulnerable to a pro-inflammatory situation, which might well contribute to the manifestation of chronic inflammatory diseases after menopause.

Most in vitro studies have demonstrated a stimulatory effect of E2 on secretion of IL-4, IL-10 and TGF-beta typically at peri-ovulatory to pregnancy levels.

Estrogens are known to modulate inflammatory processes in chronic inflammatory diseases such as rheumatoid arthritis, SLE, MS, thyroiditis, intestinal, renal and liver inflammation. However, a uniform concept for the action of estrogens has not been found for all these diseases. If B-cells are dominant in an inflammatory disease, estrogens at all levels stimulate the disease process. The opposite is the case in chronic inflammatory diseases where monocytes, macrophages, dendritic cells, T-cells, fibroblasts and neutrophils play a dominant role. In these diseases estrogens demonstrate a dual role: at low concentrations estrogens stimulate, and at high levels, estrogens inhibit the disease process.

If B-cells play a central role by means of antigen presentation, autoantibody production, and/or bystander cytokine secretion, E2 will probably speed up the outbreak of the disease in the early reproductive years.

If T-cells play an equal or more important role than B-cells, the onset of disease in a woman will be delayed because E2 inhibits T-cell autoimmunity but stimulates B-cell autoimmunity. In that situation, the onset of the disease might be shifted to the late reproductive phase or postmenopausal period.

In strictly B-cell dependant diseases the female to male preponderance can be explained by the propagating effects of estrogens. The smaller the influence of B-cells and the greater the weight of T-cells and other cells, the less evident is the sex dimorphism in chronic inflammatory diseases. Sexual dimorphism of diseases may also depend on factors other than sex hormones.

Gender and longevity

Attainment of longevity is influenced by a complex interaction between environmental, genetical and stochastical factors. Several genes are positively related to longevity. These genes influence inflammatory immune response and stress response. Polymorphisms of most genes involved in longevity have a gender-specific association. Males seem to depend on their genes to attain longevity more than females do. Females may also rely on alternative routes (such as life style, or perhaps estrogens) to attain extreme ages. This would explain why females have a higher life expectancy than males.

For instance, in Sicily, high birth rates and low iron diet (consisting mainly of grains, vegetables and fruits) were the common rule among women born around the beginning of the last century. Meat was only made available for men and not for women. This has probably attributed to the high female mortality rate in southern Italy. Also, it was suggested that women's possibility to achieve longevity was highly determined by their possession of iron sparing alleles.

In southern Europe the relatively high number of female centenarians (high F/M ratio) is much lower than in the northern part. The reason for such a gradient is presently unknown. Different rates of mortality between men and women, as well as social and anthropological differences are considered to play a role. It has been previously reported that this ratio is higher in populations where life expectancy increases as a result of economic and social reasons. Hence, a complex interaction of environmental, historical and genetic factors, differently characterizing the various parts of Italy, likely plays an important role in determining the gender-specific probability of attaining longevity.

PUGLIATTI, Sassari, Italy: Clustering of multiple sclerosis, age of onset and gender in Sardinia

Clustering, age of onset and gender are ways of finding out more about disease aetiology and age of disease induction in MS. There are different ways of clustering. Spatial clustering can disclose significant disease variation at a microgeographic level, as a possible expression of spatial variation in the distribution of disease risk factor(s).

Temporal clustering applies to incidence trends. It is to disclose significant disease variation over time, as a possible expression of temporal variation in the concentration of disease risk factor(s). Space-time clustering can disclose clues as to the period of disease induction by assessing whether individuals with MS had lived close to one another in time and space during the disease pre-onset history and if so at what age. Age of onset can disclose clues to the period of disease induction. Gender related sex-ratio changes are likely to be determined by increase in prevalence of affected women over time and can provide clues on MS risk factors.

The epidemiological interest in the island of Sardinia has to do with its high MS prevalence as compared to other areas in Italy. A first study among the present study population carried out in 1997 in the northern province of Sassari showed a prevalence rate of nearly 150 MS patients per 100,000. An a priori prevalence based spatial clustering analysis was carried out in Sassari identifying six sub-areas based on linguistic features. In this spatial clustering a Bayesian approach was used. Mapping low crude estimates in small areas for rare diseases biases the variability due to sampling or chance. The principle behind this methodology is that random variation from the estimated rates due to the small numbers are filtered out ("smoothened”). And thus a distribution of standardized prevalence rates by Maximum Likelihood Estimates is obtained. A clustering pattern in the west of the Sassari province and a west-to-east gradient appeared to be fairly evident after mapping the Bayesian estimates. Similar evidences were obtained when we mapped the prevalence by residence of MS patients at age 5-15 years. 17 out of 89 communes had a Bayesian estimated prevalence of 150 or higher per 100,000, i.e. higher than the province standardised average rate. 3 of these communes are considered "hot areas”, whereas for the remaining 14 communes only an indication for a higher risk could be given. To some extent, Sardinians are considered an isolated population with higher inbreeding rates for interior mountainous areas, which are explained by millennia of genetic drift in small but isolated populations. A founder effect was reported for MS for a village in southern Sardinia. In this village 11 cases were observed, all of them descending from 3 couples of ancestors living in the 18^th and 19^th century.

A study in northern Sardinia showed a steady rise in incidence rates between 1970 and 2008 up to 9.5 (per 100,000/year) among the overall population, 13.2 among females and 5.8 among males. These rates are 4-5 times higher than the mean national incidence rates in Italy. Mean annual adjusted incidence rates for this time interval were 6.2 for the overall population, 8.6 for women and 3.6 for men.

Space-time clustering

MS is a multifactorial disorder with a genetic complex trait. Genes are believed to interplay independently or interactively with exogenous agents to start (at some point in an individual's life) a systemic condition (non-pathological to the nervous system) which has been referred to as the ‘MS trait', or also the ‘disease waiting to happen'. This condition is characterized by expression of immunological hyperreactivity. The MS biological onset occurs the moment that immunoactive substances penetrate into the CNS after a further and more specific antigenic challenge. For clinicians their patients' MS history usually begins from the clinical onset and continues with the patient's subsequent disease course. But indeed it is the patients' pre-onset history (i.e. the ‘latent period') which is informative as to the causative agents inducing the disease. Despite the great potential of investigating the MS latent period in MS etiological research, few such studies are reported in literature.

Space-time cluster analysis is one of the approaches to be used to investigate the latent period. The concept behind space-time cluster analysis is that if the number of observed individuals that have been close in space and time prior to disease onset is higher than expected just by chance, then these individuals may have been exposed to disease risk factors. In a space-time cluster analysis performed in the province of Sassari temporal closeness was defined as having been born within two years of each other. Spatial closeness was defined as living in the same municipality. In this study the rate of observed cases vs. expected cases was significantly higher than 1 among individuals in early childhood (ages 1-3), implying that this might also be the age of susceptibility for MS.

Subgroup analysis showed a marked space-time clustering at age 1 in women, in patients that would later develop RRMS and in inhabitants of the eastern part of the province. Also, it tended to occur in patients with an early onset of the disease (under 30 years of age) and younger individuals (born after 1959). This implies that these subgroups were probably exposed to some MS risk factors in early childhood.

A space-time clustering analysis was performed from MS clinical onset backwards by a fixed number of years to disclose patterns of space-time clustering reflecting fixed latency periods between disease induction and clinical onset. No space-time clustering patterns were found in this analysis, implying that MS-specific infectious agents with fixed incubation times are not likely to be a cause for MS.

Age of onset

The mean age at onset has increased significantly in Sardinia from 26 to 31 years between 1970 and 2008. No gender effect was observed. Ages of onset of men and women were similar (30.2 vs. 29.8 years). It also appeared that the rise in age of onset was higher in women. This can be explained by efforts to improve MS diagnosis among women. Finally, a consistent increase in sex ratio's by year of birth was observed (1.7 in age group 1920-1939; 3.1 in age group 1970-1979), suggesting that more women are born susceptible for MS.

Gender

Among Sardinians evidence was found of a higher risk of MS in relatives of male probands (Carter effect).

Conclusions

In the Sardinian population evidence was found of:

- spatial clustering (prevalence based)

- space-time clustering in early childhood (1-3 years) especially in the female population, among recent cohorts of patients and at earlier age of MS onset

- steady increase of incidence over time (no temporal clustering)

- increase of female incidence cases over time

- increased sex-ratio by year of birth (more females are born susceptible to MS)

- increased age of onset over time

- increase of concentration and/or changes of risk(s) factors

- risk factor(s) that are not related to fixed latency incubation period

- genetic load could account for higher risk for MS in relatives of male patients

So it could be that recently the concentration of risk factors has changed among the population, that these factors are probably gender-related and affect the early stages of life. It could well be that the same factors also drive the changes in age of onset.

DEBOUVERIE, Nancy, France: Increasing incidence of multiple sclerosis in women in Lorraine (France)

An MS survey was carried out in the Lorraine (France) with a total population of 2,310,376 (based on National Census, 1999). Studies from several countries (including the USA, Canada, Australia and Japan) have shown an increase in female to male sex ratio in MS. This seems to be the result of a disproportional increase in the incidence of MS among women.

The objective of this study was to determine the standardized annual incidence rates in Lorraine and the secular trend over the period 1990 – 2002. During this period changes of the demographic and earlier clinical data were studied. Also a comparison was made of the course of the disease according to gender.

LORSEP cohort

The incidence rates were determined for the period 1990-2002 and the incidence cases ascertainment was based on the year of the clinical onset. All cases, with definite or probable MS according to Poser's classification were included. We examined all their neurologists' clinical records. We obtained data from the primary medical files regarding the first neurological episode, clinical course, and disability, CSF and MRI data. The data were computerized using the standardized European Database for Multiple Sclerosis (EDMUS) System. A cohort (LORSEP cohort) was constituted and included all the patients with a diagnosis of MS examined at least once (i.e. all prevalent and incident cases). Data were entered retrospectively when the patient was first seen by the neurologist before 1996. Then, data were collected prospectively after each visit to the neurologist and checked for their consistency with previous information. Individual case reports including identification and demographic data and key episodes in the course of MS were examined. Data collection was approved by the French National Commission for Data Protection and Liberties (CNIL). To date the LORSEP cohort consist data information about 3,676 MS patients (3,152 definite and 524 probable cases of MS).

Incidence rates

The annual incidence rates were calculated using log-linear Poisson modelling, age and sex standardized. The mean incidence rate during the study was 5.4/100,000 inhabitants, varying between 4.3 and 6.5. The incidence rate per year was higher in females (7.5) than in males (3.2). In this study period the incidence of MS in women increased whereas incidence rates in men did not change significantly.

One should consider the possibility that an increase in incidence rates is due to time to diagnosis from onset of MS signs or symptoms (MS was diagnosed earlier towards the end of the study period). In this case this is very unlikely because the incidence rates were based on the year at the onset of MS. Also the mean age at onset of MS increased during this period.

Secondly, the increase in incidence rates could be related to an increase in the proportion of patients with mild disability at diagnosis. In another study the disability profile of MS had not changed over a 10-year period in a population-based prevalence cohort. Similarly, a change in either the initial or subsequent disease course was not observed in the LORSEP cohort.

Thirdly, changes in the physician's behaviour should be considered as a possible reason for changing incidence rates during the study period. Evidence suggesting that treatment administered earlier in the course of disease produces greater effects might mean that physicians started treating MS earlier. In addition, the increased availability of MRI improving MS detection may explain an increase in incidence rates. On the other hand, it is possible that MS may have been underestimated during the years towards the end of the study period.

The main limitation of this cohort is that it is still relatively recent. A population-based survey avoids some problems of interpreting natural history data obtained from a hospital database.

Course of disease and gender

At the beginning of MS disability is higher in men than in women. However, from EDSS 4 no differences are observed. With multivariate models including many clinical factors no differences based on gender were seen.

SOELBERG SØRENSEN, Copenhagen, Denmark: Increase of incidence of multiple sclerosis in women: data from the Danish Multiple Sclerosis Registry

There have been many reports that the epidemiology of MS is changing. Important questions to be asked are:

- Is there a change in the incidence of MS?

- Is the change in incidence localized or universal?

- Is the change in incidence confined to women?

There is a well known but somewhat old map of the global geographical distribution of MS. It still holds true on a gross scale. The high-risk areas for MS prevalence (in Europe and Northern America) show a not fully consistent north-south gradient. The pattern mainly reflects racial differences. It is easily recognized that high-risk areas are predominantly populated with Caucasians, but climate and socio-economy also distinguish low-risk from high-risk areas. Studies from Australia, showing a prevalence of MS about half that of people of equivalent ethnicity in northern Europe, indicate significant exogenous influence. There is a theory linking MS to the Vikings spreading their genes across Europe and Northern America. MS is particularly frequent in areas where the Viking have been or where their descendants live.

Indeed, there is a genetic reference to MS. Carriers of specific HLA phenotypes (e.g. HLA-DR2 in northern European Caucasian populations and HLA-DR4 in Mediterranean populations) have a 2-4 times higher MS risk. But there is also evidence of environmental risk factors for MS in addition to genetic factors. The environmental theory is supported by the variation of MS prevalence by geographic location and the observation that migration may alter the risk of MS occurrence. Concordance rates in monozygotic twins are far less than the 100% that could be theoretically expected if MS would have a purely genetic background. Also epidemics have been reported. Clustering of MS means accumulation of cases to a density exceeding the surrounding density more than expected by chance. Pure geographical and stable clusters can indicate concentrations of people with high genetic determined susceptibility in isolates. Clusters can be time limited and disappear again. If so, it indicates changes in exposure to exogenous factors. Finally, various association studies have linked MS to environmental factors.

Regarding the increase in MS incidence, the following can be said. Repeated epidemiological surveys almost invariably show an increase in incidence and prevalence of MS. This increase is at least partly reflected by better case ascertainment. This has been regarded as the major contributor, whereas a true increase in incidence has been questioned. The problem is important as a true increase in incidence indicates a response to a stronger or more widespread exposure to environmental factors with time and can hardly be ascribed to genetic factors. Long-term registers, such as the Danish MS Registry, are needed to better understand this situation.

This Danish MS Registry was established in 1949 by Danish neurologist Kay Hillested. The registry has since 1949 collated data on virtually all cases of MS in Denmark from multiple sources. Based on the records the cases are reviewed by the staff of neurologists as to the compliance with diagnostic criteria and a number of clinical and demographic variables. Its database contains information of more than 60,000 MS patients. The registry receives notification from various sources, such as all 15 Danish Departments of Neurology, the two Danish MS rehabilitation hospitals and practising neurologists. It is linked to the National Patient Registry, founded in 1977. Since then all hospital admissions, inpatients as well as outpatients, have been kept in a national database by personal identification (a unique CPR-number), code for hospital and department, date and mode of admission (inpatient or outpatient) and discharge and ICD-code for primary and secondary diagnoses. The Danish MS Registry is also linked to other public registries.

The number of prevalent MS cases in Denmark has increased over the years. This increase is almost completely due to a rise of older MS patients (>60 years). The MS incidence in Denmark has doubled between 1960 (3.5) and 2000 (7.0). The incidence of onset in men in the period 1950-2000 has been virtually constant. This was also the case for women until 1975. Since then it has more than doubled. When looking at onset by age and decade in women (period 1950 – 2000), the incidence has increased in all age groups, except teenagers. MS incidence rates among women have increased in all age groups, but mostly in women over 40 years old.

As a result the F/M sex ratio has almost doubled from 1.3 in 1950 to 2.2 in 2000. This is in accordance with trends in sex ratio seen elsewhere (such as Canada or Italy).

The most probable explanation for the increase in sex ratio is a change of environmental factors. Some evidence of environmental risk factors for MS in addition to genetic factors is available concerning Vitamin D (and exposure to sunlight), cigarette smoking, sex hormones (and pregnancy), stress and infections (particularly EBV).

When looking closer at women, what has changed since 1975? Women have entered the labour market, they have increased cigarette smoking and age of child birth has been delayed. Other possible factors are an earlier sexual debut and the use of the anti-conception pill. There is an intriguing parallel between the increase of MS incidence in women and the average age of women when giving birth to their first child.

Conclusions

The aetiology of MS still remains elusive. Interactions between genetic and environmental factors are still only partially known. In Denmark there is a clear increase in the incidence of MS, apparently exclusively in women. This increase indicates a response to a stronger exposure to environmental factors with time. Unfortunately, these environmental factors have still not been identified.

EBERS, Oxford, UK: Life style changes and demographic patterns of multiple sclerosis: clinical evolution and gender

In order to confirm the existence of geographical patterns in MS one would ideally need a country with a homogenous, non-migratory population, distributed evenly and with a substantial north to south extent.

MS prevalence figures in France (Vukusic, 2007) not only show a North-South gradient but also a west-east gradient. This geographical distribution seems to match the amount of sunlight (UV radiation) in these areas. MS prevalence figures are lowest in areas with high amounts of sunlight (as measured in the month of February). Though this doesn't prove anything, it illustrates the remarkable importance of the geography of MS prevalence and confounding factors, such as amount of sunlight.

Familial environment

Canadian MS prevalence figures show a 400 fold difference between ½ siblings (raised apart or together) and other familial relations (e.g. spouses, step siblings or adoptees). MS prevalence is by far the highest among monozygotic female twins (MZF). This implies that the increase in MS prevalence in monozygosity is determined by the female gender. This is an important observation that tells us something about pathogenesis.

The Canadian CCPGSMS survey (30,000 MS patients) has shown a dramatic increase in sex ratio by year of birth. Sex ratio has virtually tripled over the last 2 generations as a result of an MS increase in females. Ascertainment can be ruled out as a possibility for this increase.

Lifestyle

80% of MS cases can be prevented by moving to Queensland. A number of lifestyle related risk factors for MS have been suggested include smoking, birth control pills, vaccination, hygiene, refrigeration, antibiotics, fast/processed food, women in workplace and vitamin intake. Birth control pills were introduced in the 1960's and widely used later in the decade. Sex ratio essentially excludes this factor because the change in sex ratio was well in motion long before the use of birth control pills. The same goes for the introduction of vaccination. Hygiene may be a factor for auto-immunity, as it may be in rheumatoid arthritis. Refrigeration became widespread in the 1950's. Since then stomach cancer has plummeted because of a reduction of nitrosamines in the diet. Nitrosamines were used for food preservation and their use was reduced when refrigeration became available. The relation to vitamin intake is difficult because it co-associates with many other health-conscious behaviours like exercise, amount of TV watching or smoking. But even then it is not necessary that risk factors are confined to the patients themselves. If that is so, then all case-contemporary-control studies are probably a waste of time for explaining the broad population differences.

Sun and vitamin D exposure

Sunshine was first mentioned as a possible risk factor in the 1960's by Acheson. A difference in MS prevalence between the coastal area and interior of Norway was related to climate and fish diet according to Swank. Another Norwegian study in Tromsø showed less outdoor exposure in childhood for those with MS versus controls.

An American study by Ascherio and Hollis showed that high levels of sunshine protect, but that low levels do not enable MS onset.

Norway is a discrete exception to the latitude gradient and indicates that perhaps Swank was right. Because if sunlight alone would be a determining factor, the maximum MS prevalence rate would be found in the northern part of Norway where there is no sun for 2 months in a year. Instead the highest prevalence rate is found in Oppland (interior Norway). Its prevalence rate is double that of the Finnmark rate in the north of Norway. This illustrates that one should never be looking for single explanations in complex issues. In fact, the Norwegian data provide a mechanism for a climate-diet interaction. This climate-diet mechanism was also discussed in the original Canadian twin study (NEJM, 1986).

Sexual activity

There is no clear relation between MS and sexual activity. The age of puberty has decreased and we have recently found age at menarche is slightly earlier in MS conferred to controls. However, this is a very small difference (highly significant statistically) less than 1 year. The age of first sexual experience has markedly reduced. The number of partners has soared and the age of first pregnancy has increased in western countries. The mechanism concerning MS prevalence is not clear. There is no generally coherent relation to the changes that have taken place. There is no risk for spouses attributable to having a partner/spouse with MS.

Gender-related bias

One has to take into consideration that there is a remarkable gender-related reporting bias. No better example than men saying they have had more total sexual partners than women say. Somebody is not telling the truth. This bias is reversed for familial information. For every male interposed between the source and the object we found a 10% loss in mere existence. So males say they have fewer relatives than do women. Wives of MS males report 3 times as many cases of autoimmune disease in his family as he does. This extraordinary bias accounts for reports of increased autoimmunity in MS families.

Smoking

There are multiple confirmations that smoking is more common in MS patients. But it is not clear whether this relation is primary or secondary to co-associates. Smokers are known to have different diets, BMI's, exercise patterns, outdoor activities, etc. It seems that smoking is not associated with outcome of MS.

Conclusions

Over the last 100 years many changes in lifestyle are largely female-specific. There is no lifestyle association that directly parallels change in sex ratio at this point, but there is more work to be done. Maybe sex ratio can be used as a surrogate for incidence.

It is easier to say what is not related than what is related to sex ratio. Smoking does not look promising. Also there is no support for sexual activity. Relation to menarche is positive but this is a small effect and seems to be part of an earlier maturation profile with many co-associates.

It remains unresolved but extremely important to replicate why 80% of MS should be readily preventable by going to Queensland instead of Tasmania. Climate/diet remains the main suspect, but it is unknown how and when. Nothing really eliminates vitamin D which has a very large scope with more than 1,000 genes potentially regulated by vitamin D based on VDRE distribution. Time outdoors and diet look the most viable.

EDAN, Rennes, France: Gender and disease evolution in multiple sclerosis

The relationship between disease evolution and gender is a very important issue in MS. It is a little bit disappointing that several previous studies have concluded that gender does not influence disease evolution. But there have also been some major studies that have concluded that gender does influence disease evolution, e.g on the time to reach EDSS 6, on the percentage of SPMS and on the time to use a cane from clinical onset. I looked into the existing EDMUS Rennes database for a relationship between disease evolution and gender. This database consists of 2,054 MS patients, of which 1,609 (80%) patients with RRMS and 445 (20%) with PPMS.

When comparing patients with clinical onset of MS before and after 1990, it is clear that the gender ratio has increased over time (2.0 to 2.6). Interestingly, this increase was only observed in the group of RRMS (2.2 to 3.2) and not in the group of PPMS. These figures were confirmed by results from an incidence cohort in western France in 2000 – 2001.

Concerning initial MS characteristics by gender no differences were observed in age of MS onset, presentation of initial clinical symptoms and activity of the disease (measured by 2 or more relapses in first 2 years). A clear difference was found in the percentage of RRMS in the group of females (81.8%) compared to the group of males (70.3%). Also the residual deficit after the first attack was higher in the male group than in the female group. A higher percentage of males reached EDSS 3 (76.6% vs. 65.5%) or EDSS 6 (42.5% vs. 31.7%). This could mean that the disability is worse in the male group.

Disability progression was measured by the time from clinical onset to irreciprocal EDSS 3 and EDSS 6 and by the time from EDSS 3 to EDSS 6. Time between clinical onset and EDSS 3 is shorter in men, but there are no time differences between EDSS 3 and EDSS 6. These differences in disease progression were only found in RRMS patients and not in PPMS patients.

Age at disability milestones by gender

Interestingly, the age of onset was similar in both sex groups (31.6 vs. 31.4) in the total MS group. The age of men was younger when reaching EDSS 3 (40.4 vs. 42.7) or EDSS 6 (52.6 vs. 54.9). These observations were similar in the RRMS group, meaning that disability progression is quicker in men. In the PPMS group the age of onset in males was also younger than in females, leading to a remaining younger age at EDSS 3 and EDSS 6 in men. Disability progression is similar in men and women.

It is clear that in an univariate analysis gender seems to influence disease progression. However, confounding factors have to be taken into account. Therefore a multivariate analysis using Cox-models was needed to look for an independent effect on gender, adjusted on other prognostic factors. This analysis made it clear that in the total MS population gender is a prognostic factor related to time from clinical onset to EDSS 3 and to EDSS 6, but not for the time from EDSS 3 to EDSS 6. In the RRMS group the same conclusions were made concerning the time from clinical onset to EDSS 3 and EDSS 6. Also there was a clear gender difference (M>F) in the activity of the disease during the first two years of the disease and residual deficit after the first attack. In the PPMS group not one significant prognostic factor was found related to gender, age of onset or superimposed relapses.

WINGERCHUK, Scottsdale, USA: Neuromyelitis optica (Devic). What is the role of gender?

What we know about the gender issue in Neuromyelitis Optica (NMO) (also called Devic's Disease) has been influenced by the evolution of diagnostic criteria. NMO was historically defined (since the 19^th century) as a monophasic disorder consisting of fulminant bilateral optic neuritis (ON) and myelitis occurring in close temporal association.

Diagnostic criteria

The 1999 NMO diagnostic criteria (Mayo Clinic) emphasized certain clinical and imaging factors in order to distinguish NMO from MS. These criteria advanced several concepts. One was the further recognition of associations. NMO is more common proportionally in non-white populations and is also associated with the co-existence of systematic autoimmunity (which has a female predominance). A marked increased ascertainment of NMO has been the result of better recognition of the disease (MRI, CSF and NMO-IgG serology data have been incorporated into the disease criteria) but also because of the gradual acceptance that NMO can have relapses. This has caused a shift (at least in northern America) from cases diagnosed with severe MS towards cases diagnosed with NMO. It has also caused a change in gender ratio, because most patients with relapsing disease are females.

Gender ratio

There have been some population-based studies (from Martinique and Japan), but most published gender ratio data are derived from non-operation based cohorts around the world. In Northern and Southern America F/M ratio's are generally 3:1 with the exception of Brasil (5:1 to 9:1). Asian ratio's vary from 3:1 to 9:1. Rates in Europe and South-Africa are around 3:1 to 4:1. There are also cases of familial NMO. Almost all of these cases are females (93%, 14:1), common pairs are sisters, mothers-daughters and nieces-aunts.

Gender and disease course

The disease course of NMO seems to be influenced by gender. There is a noted trend towards a decreased frequency of diagnosed monophasic cases of NMO. The increased ascertainment of relapsing disease has generated the identification of more female cases and a higher gender ratio. The F/M ratio to clinical course in relapsing NMO is around 8-9 vs. 1, meaning that 80-90% are females. In monophasic NMO the F/M ratio is 1:1 as it has been for the last 50 years. In a multivariable model the female gender is highly predictive for a relapsing course of the disease (RR = 10.0). Gender doesn't seem to influence other clinical behaviour related aspects such as severity of initial attack, recovery of initial attack, attack frequency or mortality. Published data on pregnancy and post-partum period are not available. NMO-IgG seropositive rates are similar in men and women. It is unclear whether gender influences therapeutic response. Almost all the patients in published data with respect to therapeutic response have been female, so there are little data available that tell us how males respond to therapy.

Conclusion

NMO diagnostic criteria revisions are likely to have influenced gender ratio data. F/M gender ratio is consistently higher than 3:1 worldwide, and probably higher in relapsing disease. Female gender is associated with a relapsing course and familial NMO is almost exclusively female. Gender influence on phenotype, severity and treatment response is still unclear. The specificity of NMO-IgG will probably facilitate biological investigations into the gender effect.

Clinical evolution and gender

HAAS, Berlin, Germany: Pregnancy and multiple sclerosis

MS mainly afflicts young women. In 50% MS starts between 20 and 40 years of age. The mean age at first delivery in western countries lies between 25 and 30 years. This age is still increasing. Pregnancy planning plays an important role in women with MS. Therapeutical recommendation to treat after the first exacerbation with high risk for MS plays a role. In pregnancy the developing child is a so-called semi-allograft. Immunotolerance is necessary to sustain pregnancy (and to avoid immunological abortion). The magnitude of differences within the HLA system between mother and child determines the amount of necessary immunotolerance

There is a Th1 to Th2 shift which leads to a decrease of cell-mediated autoimmunity and an increase of antibody-mediated autoimmunity. This is induced by pregnancy associated proteins, gestation hormone levels, production of corticosteroids by the placenta and insulin-like growth factor 1 beta. Pre-implantation factor (PIF) is mentioned as the main player to induce immunotolerance in pregnant women (Barnea, 2007).

Concerning pregnancy and MS a number of urgent questions need to be answered:

- What is the influence of MS on fertility, pregnancy and the developing child?

- What is the influence of pregnancy, delivery and the lactation period on the course of MS?

- What is the best immunological treatment before, during and after birth?

- What are the long term outcomes after pregnancy?

Data concerning decreased fertility in untreated female MS patients or increased spontaneous abortions rate are not available. Immunological therapy may influence fertility. The mean age of pregnancy is still rising. According to the Berlin database Musis the mean age of pregnant woman increased from 25 years in 1996 to 35 years in 2007. Also the number of first pregnancies in women above 40 years has increased. Some publications have reported an increase of relapse rate during reproduction techniques. Different reproduction techniques may have a different influence on MS exacerbation. The risks of these procedures are now under investigation by a task force. Regarding the course of pregnancy and delivery in MS a study was performed using the medical birth registry of Norway. This study gave the impression that a higher proportion of neonates from MS mothers were small for gestational age. Also more frequent induction and operative interventions during delivery were observed.

The first important study on pregnancy and MS was the PRIMS study (Pregnancy In Multiple Sclerosis), published by Confavreux in 1998. This study used prospective data from 12 European countries and a follow-up of 266 pregnancies in 251 women. The study showed a decrease of exacerbation rate during pregnancy and an increase after delivery. No changes in disability were observed during the observation period. Also no influence of epidural anaesthesia was observed on relapse rate and disease progression. There was an advantage of breast feeding.

In some patients you can observe amelioration of MS symptoms during pregnancy. That may be due to a promotion of remyelination by prolactin or repair processes by invading embryonal stem cells.

Manifestation of MS during pregnancy is rare, but in literature cases are described of Marburgs MS type and Neuromyelitis. Also a severe disease course is described in single MS cases during pregnancy, which may be attributed to an antibody-mediated subtype of MS. In these cases a pre-termination of pregnancy should be considered to stop the activity of the disease.

During delivery a lot of immunological changes occur, such as a breakdown of immunotolerance and graft-versus-host reaction. Also autoimmune diseases, such as MS, Hashimoto thyroiditis and rheumatoid arthritis, are activated. Many changing immunological factors play a role here, such as IL-12, TNF-alfa, cortisol, norepinephrin and serum 1,25 dihydroxy-vitamin D₃.

Long-term influence of pregnancy on MS has been described in 8 studies. These showed no influence of pregnancy on the long-term disease course. However, a protective effect was observed in 3 studies concerning the age of manifestation for early pregnancy. One has to consider that most studies on pregnancy and the long-term course of MS are retrospective, and that only some are prospective. This means that there has been no randomization and that the data have been collected from MS centres and are not population based. This may very well influence the observed results. For pregnancy planning after MS manifestation depends largely on the course of the disease. And pregnancies are planned more often by patients with a low EDSS.

The choice of treatment before, during and after pregnancy is determined by the activity of the disease and fertility, course of pregnancy, developing child and breast feeding.

Use of corticosteroids prior to pregnancy seems to have no negative impact on fertility. Corticosteroids pulse therapy can cause menstrual disorders due to an effect of the gonadotropin releasing hormone. During pregnancy continuous use of corticosteroids > 25 mg/day within the first trimester can be harmful for the developing child. In rheumatoid arthritis it has shown to cause a 3.5 fold risk increase of cleft palate. Low oral dosages are safe and also used in fertilization therapy. Corticosteroids in advanced pregnancy may increase risk of hypertension, gestational diabetes, venous thrombosis and small for gestational–age babies. Experience in polyarthritis has shown no negative long-term influence in ten years data.

Corticosteroids cross the placenta, but only ten percent reach the foetus. A short course of prednisolone is preferred for MS exacerbations during pregnancy. Dexamethasone and betamethasone cause danger for the respiratory health of newborns because they cross the placenta in equal maternal and fetal concentrations.

During breast feeding corticosteroid pulse therapy is possible. Breast milk should be withdrawn 4 hours after infusion in order to reduce the risk of exposure to corticosteroids.Corticosteroid levels in breast milk may have a neuropsychological effect on the baby during its early phases of development. No data exist for breast fed babies of prednisolone pulse treated mothers.

INFb (1a and 1b) for MS doesn't decrease fertility, but increases the rate of spontaneous abortion. Birth defects have not been observed though birth weight is lower. INFb passes through to breast milk and may be harmful for babies.

GA has no effect on fertility, abortion rate or birth defects. It cannot pass through to breast milk, though particles can and may be harmful.

In animal studies natalizumab has shown a fertility decrease in high dosages only. Human data are not available. It increased the rate of spontaneous abortion in one animal study, while human data collection is ongoing. Natalizumab can cause haematological changes (especially thrombocytopenia) in young animals. This is not yet observed in humans. Low concentrations can be found in breast milk that may be harmful for babies.

Azathioprine has been shown to cause a dose-dependant decrease in fertility, an increased abortion rate, lower birth weight, congenital abnormalities (in transplantation patients). It may be harmful during breast feeding.

Use of mitoxantrone decreases fertility and provides a risk for ovarian failure. It increases abortion rate and has been observed to cause congenital abnormalities in animal studies. Human data are not available, but due to the mode of action of the molecule a high risk for abnormalities in humans is suspected. Breast feeding is harmful.

Use of immunoglobulins (though not licensed for MS yet) has no influence on fertility, may even decrease abortion rate and provides no risks for breast fed babies. Congenital abnormalities have not been observed.

Female patients with a clinical isolated syndrome (CIS) or MS with a low exacerbation rate (≤ 1 per year) and an EDSS ≤ 2.5 who wish to become pregnant are advised to start immunological therapy after delivery.

Female patients with a low exacerbation rate and an EDSS ≤ 2.5 who use immunomodulatory therapy with INFb or GA and wish to become pregnant, are advised to stop their treatment or go on with IVIG if possible.

Women with a high exacerbation rate (>1 per year) and an EDSS ≤ 2.5, who have been stable since their immunomodulatory therapy with INFb or GA and wish to become pregnant, are advised to get pregnant and then stop their therapy.

Female patients with a high exacerbation rate, an EDSS ≥ 3.0 with severe disabling exacerbations, who have been stable since their immunomodulatory therapy with INFb or GA and wish to become pregnant, are advised to continue their therapy during pregnancy.

After delivery no therapy or IVIG should be given when the patient is breast feeding. When the patient is not breast feeding INFb or GA is advised. In case of a malign course of MS after delivery, natalizumab or mitoxantrone should be used.

In conclusion, the treatment decision during and after pregnancy depends on the activity of the disease. The need for corticosteroid pulse therapy reflects this activity. Drugs that cross the placenta may harm the unborn and should be avoided. A shared decision with the informed patient should be made.

PALACE, Oxford, UK: Leber's disease and gender

The reason to look at Leber's Hereditary Optic Neuropathy (LHON) is that female patients have the highest known risk for MS. Having such a high risk, may teach us something that is more generalizable to MS patients. But first some important questions need to be addressed to be sure about the relevance for MS.

LHON is maternally inherited and causes acute or sub-acute severe bilateral visual loss. Of interest is that males are characteristically affected. LHON is caused by three primary mitochondrial DNA point mutations which encode subunits of the complex 1 respiratory chain.

A sample size of 46 cases of Leber's were detected from 23 publications going back to 1964 and was used to address the following questions:

1. Do Leber's and MS co-occur by chance alone?

There are a couple of Leber's cohort studies in which patients were screened for MS. The observed frequency of MS in Leber's is around 5%, which implies a 50 times higher MS risk than in a normal population. This combination of Leber's and MS, also known as Harding's Disease (HD), is still very rare. In MS cohorts Leber's mutations are rarely found.

2. Is the non-ON in Harding's Disease actually MS or could it be mitochondrial damage?

Ways of looking at this question is by gender and mutation distributions, MRI features, pathology and clinical characteristics (natural history and treatment effects). The female predominance in HD is very similar to what is seen in MS, suggesting that this could well be MS. So females with LHON have a much higher risk for MS, even though their penetrance for LHON alone is much lower.

Published MRI scans of HD generally look similar to MS, though minor differences occur. These are mostly smaller lesions (< 5mm) with elongated tube shapes and indistinct margins. Also there is a lack of atrophy, juxtacortical lesions are common, there is less Gd enhancement and black holes are rare.

Pathology shows that mitochondrial defects in Leber's without MS do not only occur in the optic nerve itself, but also in the brain tissue causing diffuse microscopic lesions. It is known that mitochondrial defects can cause white matter changes in the brain which may not look MS like. In patients with HD the CSF shows positive OCB's in 70-81% of cases. The clinical picture is perhaps the most convincing. For the ON in those with isolated Leber's and those with HD appears to be very similar. Generally the ON is severe and without recovery. Occasionally there are case reports of improvements (usually temporary) with or without the use of corticosteroids. The visual outcome is slightly worse in patients with HD. However, we see a different clinical picture with regard to non-ON patterns in HD patients. Usually after ON they present a course of classical MS. They have relapses that spontaneously recover, relapses that respond to corticosteroids and some develop into a typical progressive MS course (either after relapses or without).

There is evidence to support that ON is a mitochondrial pathology and that the non-ON brain disease is likely to be MS. LHON patients have a 50 times higher risk of developing MS. There is a marked gender effect showing that LHON females have a much larger MS risk (F 45% and M 3%), but a lower penetrance for LHON alone. A possible explanation may be that females have a higher threshold of getting LHON, and that the ones that pass this threshold have a bigger gene effect.

Whether mitochondrial factors are responsible for the gender effect found in normal MS remains a question. Also it may be that the unknown MS gender effect (related to environment or hormones) is responsible for an even larger gender bias in Leber's.

FAZEKAS, Graz, Austria: Gender differences in MRI, studies on multiple sclerosis

With gender playing such an important role in MS, it is interesting to establish if there is any evidence that gender independently influences brain changes as displayed by MRI.

In a publication by Weatherby in 2000 a positive relationship between gender and Gd-enhancing lesions was described. In a small study population of 29 RRMS and 21 SPMS patients women were found to have a higher relative risk on Gd+ lesions. In 1999 Bansil et al. reported a positive association between high oestrogen/progesteron levels in women and more Gd-enhancing lesions. From a cross-sectional evaluation of gender differences in MRI characteristics of brain lesions (413 consecutive outpatients with MS of which 266 RRMS and 147 SPMS patients), Pozzili (2003) concluded that women with MS are more prone to develop inflammatory lesions, and that men with MS show more destructive lesions. From a study on the role of sex hormones on the modulation of brain damage in MS, Tomassini (2005) confirmed the earlier reported difference in Gd+ lesions between men and women (F>M), although no gender effect on lesion load could be established. The highest lesion number was found in women with abnormally low testosterone levels.

A large amount of data is available from placebo treated MS patients in randomized trials that can be used to address the question of gender differences in MRI.

From 17 RCT's a total of 1,328 placebo patients (189 CIS, 602 RRMS and 537 SPMS) were used to predict Gd-enhancement status in MS. Overall, 47% of these patients showed Gd-enhancement. Univariate and multiple logistic regression analyses were performed on these populations. The model was validated in a closed data set of 848 patients. It turned out that burden of disease, disease course, disease duration and age at disease onset were the most predictive variables for Gd-enhancement. This was confirmed by a closed data set.

From the same data set an analysis was made on T2 lesions burden in MS. A sample of 1,312 placebo patients (176 CIS, 463 RRMS, 673 SPMS) were pooled from 11 RCT's. A univariate analysis for demographic, clinical and MRI variables was made to establish their contribution to the burden of T2 lesions. Most important predictors of disease burden were disease course, age at onset or disease duration, EDSS and enhancement status.

Hypodensity in T1 lesions is pathologically correlated to the severity of brain destruction. An analysis of 138 MS patients was performed by the Department of Neurology of the Vrije Universiteit Amsterdam to determine if clinical MS characteristics are related to T1 lesion volume. In patients with RRMS or SPMS no statistically significant differences in T1 lesion volume or T1/T2 ratio were noted between male or female patients. In patients with PPMS a slight trend towards a higher T1 lesion volume and a higher T1/T2 ratio (0.22 vs. 0.08; p= 0.02) was shown in male patients compared to female patients.

In a study of brain atrophy and lesion load in a large MS population (597 patients), gender was not included in any of the models as a variable in the correlation between MRI and clinical data.

In conclusion it can be stated that available data do not provide compelling evidence for an independent influence of gender on morphologic brain changes as displayed by conventional MRI in CIS and relapsing MS with or without chronic progression.

TINTORé, Barcelona, Spain: Early onset multiple sclerosis. The role of gender

One way of looking at the role of gender in early onset of MS is to study literature about patient cohorts younger than 16 years with different clinical pictures of inflammatory demyelinating diseases, such as ADEM, MS and NMO. A second way is by looking at data about onset of Clinically Isolated Syndrome (CIS).

When looking at these different clinical pictures of inflammatory demyelinating diseases one has to take into account their distinct differences with MS. ADEM usually occurs in childhood (< 10 years) and has a high correlation with viral infection, encephalopathy and seizures. There are also morphologic differences between ADEM and MS. When looking at ADEM and gender the percentage of male patients is about 50-60%. Therefore one can conclude that in ADEM there is a slight male predominance (or that both genders are equally affected).

According to the International Paediatric MS study, MS starts in childhood or adolescence in 2-5% of MS patients. Within this age group (especially in the early affected cases) there is an overlap with the clinical and MRI presentation of ADEM in 40% of cases. Lesions in young MS patients may be bigger and also atrophy can occur. When looking at MS (in children and adolescents) and gender one might conclude that there is a difference in gender distribution with increasing age of onset. Below 12 years F/M ratio is 1:1, although some studies suggest that in children with MS less than 10 years old males are overrepresented. In puberty (age group of 12 years and above) there is a clear female predominance of around 3:1. This female predominance also exists in paediatric NMO.

The Barcelona CIS inception cohort study is a longitudinal study that started in 1995. CIS patients aged below 50 years enter the study within 3 months after CIS presentation and are regularly followed-up every 3 to 6 months. In this study no gender difference is observed in the mean age of onset. Female predominance in age of onset is highest in the adolescent group. Young CIS patients convert more frequently to CDMS, but this is not associated with gender. Also it is clear that CIS sex ratio has not changed over time (period 1995 – 2008). However, the age of onset of CIS has steadily increased during this period. This rise in age of onset had not been affected by gender.

Conclusion

In MS there is a female predominance during the reproductive ages (especially in puberty). Males seem to be overrepresented at very young ages (below 10 years old). The Barcelona CIS cohort does not show an increase in F/M ratio over the last decades, but does show an increase in age of onset over time, which is not affected by gender.

TOMASSINI, Oxford, UK: Sex hormones, brain damage and clinical course of multiple sclerosis

Gender influences the course of MS, both across and within types of MS. Motor symptoms (especially at MS onset) and cognitive impairment are more frequent in men than in women, whereas sensory symptoms (e.g. range of pain conditions) are more frequent in women than in men. Also gender influences clinical characteristics in MS. For example, within sensory symptoms men suffer more frequently from spasticity-related pain, whereas women suffer more from central pain, trigeminal neuralgia, peripheral neuropathic pain and nociceptive pain.

Concerning brain damage in MS gender differences are observed. Men are prone to develop less inflammatory but more destructive lesions than women suggesting a modulation of MS pathological changes by gender. This gender difference is also observed in intracortical inflammatory lesions (ICL's). ICL's occur more frequently in men (79% vs. 51%) and men run a risk of having ICL's that is almost 4 times higher than in women. However, we should also take into account that in MS there is a gender influence on atrophy rate in white matter tracts. Some parts of the brain seem to be more susceptible than others, especially the cingulate and the inferior longitudinal fasciculus. This suggests that gender differences in regional brain structure may affect the development of brain atrophy. And indeed there are gender differences in normal white matter brain structure. When looking at the microstructure of the corpus callosum, men have thicker myelinated fibres than women and/or the interfibre space in men is decreased compared to women (Westerhausen, 2003).

Gender differences also exist in the normal grey matter brain structure. When looking at maps of differences between the sexes in thickness of grey matter, it is apparent that men have a thicker grey matter in the dorsomedial regions, whereas women have a thicker grey matter in the temporal lobe. This is in accordance with gender related differences found in the regions that show brain atrophy.

So we could say that both MS damage and normal brain structure may contribute to explain gender difference in MS clinical characteristics. And the main actors for these differences could be sex hormones. These may be responsible for gender differences in brain damage and the clinical characteristics of MS.

MRI disease activity is associated with sex hormone fluctuations during the luteal phase of the menstrual cycle. There is a positive correlation between PEL (= Progesterone (pg/mL)/Estradiol (pg/mL) x 0.01 during the luteal phase) and the number and volume of Gd-enhancing lesions (Pozzilli, 1999). This has been confirmed by a number of studies that have investigated the relationship between PEL fluctuations and clinical MS activity. These studies found that the premenstrual period triggers relapses in a subgroup of women with MS (approximately 40%) (Zorgdrager, 1998) and that the relapse rate declines during pregnancy, especially in the third trimester (Confavreux, 1998).

Sex hormones are not only of importance in view of relapse rate, but also concerning the severity of symptoms. Hormonal changes preceding menstruation may worsen MS symptoms in a subgroup of women with RRMS (43%) (Zorgdrager, 1998). Interestingly, from the group that did not experience an influence of the premenstrual period on the severity of MS symptoms, a significantly high proportion was using oral contraceptives.

Estrogens have transcriptional effects through direct and indirect genomic mechanisms. Transcriptional effects on enzyme production mediate oestrogen actions on neurotransmitter systems (Ach, 5HT, Glu) and regulation of neuronal apoptosis. At high concentrations, estrogens exert antioxidant effects through non-genomic mechanisms. Neurotrophins (NGF, BDNF, NT 3, 4-5) and estrogens may influence each other's actions by cross-coupling of their signalling pathways that lead to induction of the same set of genes involved in neurite growth and differentiation. In the damaged brain there is a switch-back to the developmental pattern (i.e. regulation of estrogen receptor by neurotrophins) resulting in re-expression of growth-promoting properties of estrogen receptors. This switch is mediated by enhanced neurothrophin-sensitivity during damage. Estrogens also have several immunoregulatory effects. Following moderate traumatic brain injury estrogens are deleterious in females, but protective in males (Attella, 1987).

Progesterone has neuroprotective and pro-myelinating effects in EAE. Progesterone administered a week before EAE induction produces a delay of the disease onset and reduction of clinical severity. It results in less demyelination, recovery of myelin proteins and preservation of neuronal integrity (Garay, 2007). Progesterone increases the vulnerability of neurons to apoptotic injury in EAE (Hoffman, 2001).

Endogenous testosterone levels have a preventive effect on MS relapses and demyelination. Only in male SJL mice that were castrated prior to immunization, clinical EAE relapses occurred. This can be explained by the observation that widespread demyelination occurred in castrated mice versus only sparse demyelination in control mice (Bebo, 1998). When administering (exogenous) testosterone to female mice before EAE induction the clinical severity of EAE is reduced through an enhanced production of IL-10 (Dalal, 1997). Testosterone amplifies excitotoxic damage of cultured oligodendrocytes (Caruso, 2004). MS women with low testosterone levels (2SD below the control mean) had higher number of enhancing lesions than MS women with normal testosterone levels (Tomassini, 2005).

In a survey among 68 RRMS patients receiving INFb it was found that interaction between gender and brain damage influences the clinical response to treatment. At baseline greater tissue loss is associated with a higher risk of developing disability during follow-up. One year after the start of IFNb therapy male gender is associated with a higher risk of developing disability during the subsequent five years.

This suggests that gender may reflect brain damage characteristics (i.e. inflammation or degeneration) on which therapy can or cannot be effective. Or gender may also influence the clinical response to therapy beyond its effects on brain pathology. Perhaps there is a gender difference in brain plasticity (damage repair).

In order to establish the effect of sex hormones in MS treatment a longitudinal study (6 months pre and 12 months during testosterone treatment) was performed with 10 clinically defined RRMS men (18-65 years old, EDSS ≤ 5.0). The PASAT score gradually improved during the last 6 months of testosterone treatment. Testosterone treatment produced a 67% reduction in the rate of brain atrophy (Sicotte, 2007). This suggests that testosterone treatment improves cognitive performance and slows brain atrophy in men with MS.

In a cross-over study with oral estriol (6-month no treatment/6-month treatment/6-month no treatment/6-month treatment) with 10 clinically defined MS women (6 RRMS, 4 SPMS) the PASAT score improved in RRMS women after 6 months of estriol treatment (Sicotte, 2002). This suggests that estriol treatment is associated with improved cognitive performance in women with MS.

Gender affects the clinical manifestations as well as the characteristics of brain damage in MS. Sex hormones may be responsible for gender differences in clinical and MRI brain injury characteristics by modulating mechanisms of damage and tissue repair. Gender is an independent predictor of clinical evolution during therapy. This could be due to its close relationship with damage characteristics and might suggest a gender-based approach in therapy choice. Sex hormone therapy has the potential of modulating brain damage directly and/or indirectly through the enhancement of brain plasticity in MS.

Next step is a study of the effects of sex hormones on cognitive decline and brain atrophy in MS among 150 normally menstruating RRMS women (18-40 years old, EDSS ≤ 4.0). The study population will be divided into three treatment groups of 50 patients each. One group will receive IFNb only, a second group will receive IFNb + 20 mcg ethinylestradiole, desogestrel 150 mcg and a third group will receive IFNb + 40 mcg ethinylestradiole, desogestrel 125 mcg. This 2 year study will evaluate the changes in brain atrophy and cognitive performance over a two-year treatment period (control group vs. hormone groups).

EIKELENBOOM, Amsterdam, Netherlands: Sex differences in pro-inflammatory cytokines and Vitamin D

MS is thought to be an autoimmune disease with inflammatory and neurodegenerative characteristics. It is thought that the kind that starts with a relapsing remitting course of MS, in which the dominant feature is inflammation, progresses with a more neurodegenerative injury, seen as SPMS. In PPMS there is less inflammation and more degeneration, as can be shown on MRI and clinical tests.

The predominance of females is remarkable in this disease. As in many autoimmune diseases females are more susceptible, they have a milder disease course of MS and they seem to have a better reaction in RRMS to IFNb treatment. There is growing evidence that MS incidence is increasing, especially in females.

Furthermore, pregnancy has a protective effect in animals and humans with an increase of exacerbation rate post partum. One study has shown the short-term reduction of incidence in females that use oral contraceptives.

These gender differences are suggested to be caused by either sex hormones, genetics, immune bias or environment. In all these cases cytokines seem to play a key role.

Cytokines are of utmost importance in MS. The most used hypothesis for the pathogenesis of MS is the dysbalance of the regulatory Th1 and Th2 cells, as well as Treg cells. Th1 cells secrete pro-inflammatory cytokines (such as IL-2, TNFa and IFNb), while Th2 cells secrete anti-inflammatory cytokines (such as IL-4, IL-10 and IL- 13). These cytokines can be influenced by hormones. We know that females have a more pro-inflammatory Th1 response to infection. Estrogen does not merely have a stimulating effect. Low levels of estrogen enhance immune activity, while high levels inhibit immune activity. In addition cytokine production is enhanced by estrogens, which has been shown in an EAE model, but the pro-inflammatory response was decreased. In female MS patients estrogen treatment shifts towards Th2 response and increases the number of Gd+ lesions as visible on MRI. From previous studies we know that females respond with more pro-inflammatory cytokine INFg and less anti-inflammatory cytokine IL-5 than males when stimulated by various myelin compounds (MBP, MOG, PLP). Gender differences were not observed in pro-inflammatory cytokine TNFa and the anti-inflammatory cytokine IL-10, with the exception of one study that reported that males with RRMS have higher levels of TNFa.

From our own data analysis of 124 MS patients and 34 controls it was concluded that during the progressive phases of MS pro-inflammatory cytokine levels (TNFa, INFg, IL-2) are higher in females. No gender differences were found with respect to anti-inflammatory cytokine levels (IL-4, IL-10, IL-13).

We know that there is seasonal variation in immune measurements. One of the reasons that cytokines are influenced by season is probably by way of sunlight, and therefore by vitamin D levels. Vitamin D₃ can be raised by dietary intake (especially fatty fish) and UVB radiation on the skin. In the human body Vitamin D₃ is hydroxylated in the liver into 25-hydroxyvitamin D₃ and in the kidney into 1a,25-dihydroxyvitamin D₃, which is the active form. This active metabolite influences intestines, bones, tumour microenvironment and immune cells via vitamin D-receptors. It had been shown that vitamin D is preventive for EAE in animals. In a large military cohort study in the US 25-hydroxyvitamin D₃ was protective for MS. Beyond that there is a strong relation between disability and vitamin D status in MS. Vitamin D downregulates Th1 response and enhances functionality of Treg cells. In a remarkable animal study (2005) dietary vitamin D decreased EAE severity in female but not in male mice as well as the onset of EAE.

In a prospective vitamin D study among a MS cohort of 103 MS patients and 110 healthy controls in the Dutch MS centre, no differences were observed between MS patients and controls. When stratified for gender, females had higher levels of vitamin D, which were inversely correlated to MS risk. This effect was more outspoken in winter than in summer. Also a negative correlation was found between 25-hydroxyvitamin D levels and the clinical scale of neurological impairment in MS (EDSS). This was only seen in females.

In conclusion, gender differences exist in different phases of the disease. When looking at the level of cytokines, females seem to have higher levels of pro-inflammatory cytokines than males in the progressive phase of the disease. Also females have higher levels of vitamin D, which are inversely correlated to MS risk and disability.

Genetics

KNUDSEN, Oslo, Norway: Gender-bias in auto-immune diseases. X-chromosome inactivation in women with multiple sclerosis

Female predominance in autoimmune diseases

The majority of patients in autoimmune diseases are females. F/M ratio's for autoimmune diseases range from 10:1 in primary biliary cirrhosis to 3:1 in MS. Within this range are SLE (9:1), autoimmune thyroid diseases (8:1) and RA (4:1). There are several factors that are thought to influence gender bias. These include hormones (estrogens, androgens, prolactin), pregnancy (immunological challenge), fetal microchimerism (presence of allogenic male fetal cells after pregnancy), environmental factors (gender specific, with genetic and epigenetic interactions) and the X-chromosome.

X-chromosome and X-chromosome inactivation

There are several considerations why the X-chromosome can be regarded as a contributing factor to female predominance in autoimmune diseases. Several immune related genes are linked to the X-chromosome (e.g. FOXP3, SH2D1A and TNFSF5). A number of studies have shown that X-chromosome abnormalities can lead to autoimmunity (e.g. in Monosomy and Turner Syndrome). X-chromosome inactivation (XX versus XY) in females starts early in the embryonic development, when both X-chromosomes (paternal and maternal) are still active. X-inactivation occurs randomly either in the maternal or paternal X-chromosome. In some females this can lead to a skewed distribution of active X-chromosome, of which most are either paternal or maternal. In most females active paternal and maternal X-chromosomes are equally distributed. In some females a skewed X-inactivation can occur by selection. Females become more skewed as they grow older. This development starts around the age of 60 and continues from then on (Kristiansen, 2002). Skewed X-inactivation may sometimes occur in females affected by X-linked disorders, such as Heamophilia and Duchenne's muscular dystrophy. The expression of Haemophilia depends on the amount of gene product that is delivered by mutated Haemophilia genes on the active X-chromosome. This happens only when the distribution of X-chromosomes is unfavourably skewed with Haemophilia affected genes. A similar effect was shown in the case of a monozygotic female twin pair discordant for Duchenne's muscular atrophy. Both twins had a mutation on the X-chromosome, but only one of them had the disease. Analysis showed that the affected twin had an unfavourably skewed X-inactivation.

So females are mosaics as a result of X-inactivation. Consequences are that females are protected against some X-linked disorders and this might also explain why females live longer than males. But it might also influence the female predominance in some autoimmune disorders. Stewart (1998) formulated the following hypothesis on the relationship between X-inactivation and autoimmunity. When skewed X-inactivation occurs in the thymus, T-cells are educated to be tolerant only to self-antigens encoded by one of the two X-chromosomes. This results in T-cell reaction against self-antigens encoded by the other X-chromosome encountered in the peripheral tissues.

X-inactivation studies in autoimmunity have shown an increased frequency of skewed X-inactivation in blood. In systemic sclerosis 45-49% of female patients had skewed X-inactivation versus 2-8% in controls (Özbalkan, 2005 and Uz, 2007). In autoimmune thyroid diseases 32-34% of female patients had skewed X-inactivation versus 2-8% in controls (Brix, 2005 / Özcelik, 2006). No differences in X-inactivation have been observed in the blood of female patients with primary biliary cirrhosis (Miozzo, 2007) or with SLE, DM type 1, juvenile arthritis and MS (Chitnis, 2000). However, the study group of MS patients was only small sized.

X-chromosome inactivation in MS

Within a large study population of 568 females with MS (Norway, Sweden and Denmark, age 22 – 75 years, median age 43 years) and 132 controls (Norway, age 19 – 65 years, median age 40 years) skewed X-inactivation was observed among 17% of female MS patients versus 11% of controls (Knudsen, 2007). However, this was not a significant difference. No differences were found in the degree of skewing among different age groups. Between clinical subgroups (RRMS, SPMS and PPMS) a difference was found in the degree of skewing. The degree of skewing was significantly higher in the progressive (SPMS / PPMS) group compared to the relapsing (RRMS) group.

Conclusion

Skewed X-inactivation is associated with some autoimmune diseases. X-inactivation does not seem to explain the female predominance observed in MS in general. Results from the Norwegian study may indicate different genetic influences of the remitting and progressive phase of MS.

ACHIRON, Tel-Hashomer, Israel: Gene expression in relation to gender in multiple sclerosis

Many autoimmune diseases have an increased frequency of females and this is also the case in MS. Central question in this presentation is whether there is a gender variation in disease patterns of patients with RRMS. To answer this question patient data and gene expression data were reviewed from the computerized database of the MS Centre in Sheba. First disease variables were identified that showed differences between females and males. These variables were then analysed for gene expression in order to better understand the biological mechanisms involved. The study sample included 2,012 RRMS patients (1,387 females and 625 males). The following MS variables were taken into the analysis: age of onset, time to 2^nd relapse (i.e. conversion to definite MS), relapse rate (5 years), EDSS (5 years), time to EDSS 4.0 and 6.0, time to EDSS 4.0 and 6.0 with age of onset after 40 years of age, survival, cognition, happiness, corpus callosum volume and cortical thickness.

Age of onset

The percentage of female and male MS patients in each 5 years age group is similar to the F/M ratio that we are already familiar with. At young age (10 years) and above 60 years the F/M ratio is 1.0. In between F/M ratio rises to 2.8 (around 20 years old) and gradually decreases in older age groups. This possibly suggests an estrogen effect. When looking at the age of onset by correcting for the increased female prevalence and valuating the percentage of males and females in each 5 years age group in reference to the total group of the same gender, no differences in age of onset exist between males and females. And so this variable was rejected for further gene expression analysis.

Time to 2^nd relapse (conversion to definite MS)

Females convert earlier to definite MS than males. This significant gender difference is already observed after approximately 4 months and continues during the following 10 years. So this variable was approved for further gene expression analysis. When looking at the micro-array patterns 345 genes were found that correlate with the time to 2^nd relapse. From these, 24 genes were identified that are differentially expressed in females and males and gender related molecular pathways were reconstructed that are related to the type of 2^nd relapse. These pathways result in increased inflammation and/or impaired apoptosis that shorten the time to 2^nd relapse in females.

Relapse rate

Females tend to show a higher relapse rate in the first 5 years from onset of the disease. However, this difference is not significant and so this variable was rejected for further gene expression analysis.

EDSS

EDSS progression in the study population was identical for females and males in the study group. However, the number of patients that were followed-up for more than 25 years was relatively small. Also when analysis was concentrated around the first 5 years from onset, no significant differences between genders were found in EDSS development related to disease duration.

EDSS progression - Time to EDSS 4.0 and 6.0 (and age of onset)

Time to reach EDSS 4.0 and 6.0 were similar for males and females. But when combining the EDSS progression with age of onset, a gender related difference was found in patients over 40 years old to reach EDSS 6.0. No differences were found in time to reach EDSS 4.0 in this group. Females with onset after 40 years old progress more rapidly to EDSS 6.0 than males, suggesting the importance of early treatment. And so this variable was approved for further gene expression analysis. When looking at the micro-array patterns 1,300 gender related genes were found. These genes corresponded with several significant enriched pathways, including androgen and estrogen metabolism, calcium signalling and ephrin receptor signalling. 69 of these genes were related to the X-chromosome and 13 were related to the Y-chromosome. From these 1,300 genes, 102 were found to correlate with time to EDSS in RRMS patients. Next step was to analyse gender-related molecular pathways related to time to reach EDSS 6.0. Most of these pathways were aimed at T-cell clonal expansion, inhibition of apoptosis and neurodegeneration. Additional pathways were identified that contribute to shortening of time to reach EDSS 6.0 later in life.

Survival

Within the study group, no gender related differences were found with regard to survival and MS. And so this variable was rejected for further gene expression analysis.

Cognition

When analysing the data from 297 RRMS patients who participated to a computerized cognitive assessment (called Mindstream), no gender related differences were found in the normal cognitive test score nor in any of the major cognition domains (memory, motor skills, verbal function, visual spatial function, executive function, information processing speed and attention). However, two differences were found between females and males. These were related to verbal memory (better in female patients) and the time to make a first move within the Catch Game (i.e. motor function test, in which males performed much better). The implications of these outcomes need further consideration. This variable was rejected for further gene expression analysis.

Happiness, corpus callosum volume

Seeing that no differences were observed concerning degree of happiness and corpus callosum volume, these variables were rejected for further gene expression analysis.

Conclusion

Changes in genes appear to be related to the early disease activity (time to 2^nd relapse), particularly to increased inflammation and impaired apoptosis. Other gene changes are linked to T-cell clonal expansion and inhibition of apoptosis related to later stage of disease activity and correlate with the time to develop significant disability (EDSS 6.0). And so, gender dimorphic gene expression profiles are not only influential in the prevalence of MS between males and females, but are also related to clinical gender-related differences within MS.

Environmental factors and gender

COOK, Newark, USA: Environmental factors in multiple sclerosis

There is general consensus about the concept of MS risk factors. These MS risk factors are genetic, epigenetic or environmental. All of these factors interact and stimulate the pathogenesis of MS.

With respect to genetic factors there have been several genetic epidemiology studies. Also some specific immunoregulatory genes have been identified (such as HLA-DR15/DQ6, IL-2RA and IL-7R). CNS structural genes (such as GD3S/GD3) are a new gene polymorphism that also have immunoregulatory effects. These genes may have epigenetic effects as well as environmental consequences.

Epigenetics involves modification of gene expression and function (e.g. histone acetylation or DNA methylation), which is passed on from generation to generation.

Environmental factors include geography, climate, migration and incidence changes (sex ratio's and month of birth). There are several hypotheses on the relation to MS, including hygiene (infection, EBV), smoking, vitamin D status, diet and estrogens.

CNS structural genes - GD3S/GD3

The genomic structure of the human GD3 synthase gene (ST8SIA1) showing the location of the four SNP's (rs4762896, rs2041906, rs1558793 and rs704219) is associated with the risk of developing MS. Possible mechanisms of GD3S/GD3 induced CNS pathology are:

- Activated microglia releasing GD3 cause oligodendrocyte apoptosis. This is relatively specific for GD3 as compared to other gangliosides;

- GD3 can possibly lead to CNS degeneration. In GD3 knockout mice a reduced neural regeneration occurred, though this didn't lead to neurological deficits;

- Antibodies to gangliosides are found in PNS disorders and (less robustly) in MS;

- GD3 is involved in T-cell transendothelial migration as well as T-cell proliferation, NK cell activity and cytokine production. So in addition to their strong structural involvement in the CNS, perhaps these genes are modulating factors.

- Gangliosides have receptor functions;

- GD3 is rich in glia and plasma membranes (particularly in oligodendrocytes and their precursors, astrocytes, microglia) myelin and neurons. However, their functions are not well known.

There is an increased risk of MS in smokers versus controls. There is a dose effect (years of smoking, active smoking, packs per day) and perhaps even a passive smoking effect. Comments are that not all case control studies show a clearly significant risk from smoking. Smoking could be a surrogate for other factors or risky behaviour. People with mild or subclinical pre-existing MS could be more likely to smoke. There could be genes involved that predispose to smoking. And there are potential problems with recall bias, onset versus diagnosis and unpublished negative studies.

Regarding the specificity of smoking and MS, smoking may contribute to other autoimmune diseases that are not MS related (such as SLE, RA, Graves, primary biliary cirrhosis) and it may trigger or aggravate MS, RA and SLE. In a cohort study it was found that smoking may worsen MS (hazard ratio 3.6), by converting RRMS to SPMS and by enhancing brain atrophy. This is probably the result of a secondary effect. There is a environmental-genetic effect of smoking. There is a dose-dependent increase in citrullinated peptide antibodies and disease severity and risk in RA. Smokers with HLA-DRB1 (SE restricted) versus non-HLA-DRB1 non-smokers have a relative RA risk of 21. The same applies in SLE where smoking may induce double strand DNA antibodies. One study reported increased in EBV VCA antibodies in healthy female smokers versus non-smokers, but this study has not been substantiated by other studies.

Smoking can have an immunomodulatory effect on T-cells and B-cells, increasing apoptosis, inflammatory cytokines, chemokines, IL-6, ICAM-1, E selectin, and MMP's.

There are receptors for 1,25 dihydroxy-vitamin D in many tissues. Vitamin D controls more than 200 genes (directly or indirectly) and decreases normal and cancer cell proliferation. It is a potent immunomodulator, suppressing antibody production, T-cell proliferation, increasing Th2 phenotype and decreasing pro-inflammatory cytokines. Vitamin D blocks and reverses EAE, increases IL-4, IL-10, TGF-β Tregs. Deficiency of vitamin D may predispose to autoimmune disorders (MS, Crohn's Disease, RA, DM type 1). There have been a number of epidemiological studies (both observational, case control and cohort studies) that have investigated the relation between vitamin D deficiency and MS. This subject will not be elaborated now.

Bach (NEJM, 2002) was the first to mention the hygiene hypothesis in neurology. He pointed out that as infectious diseases have decreased, particularly since the mid-1960's when vaccines for certain diseases were introduced and hygiene improved, we've seen an increase in some autoimmune diseases such as DM type 1. This may also apply to MS seeing the evidence that it has increased particularly in women.

There have been a number of findings that suggest an environmental explanation for the recent changes in MS incidence. These findings include an increasing F/M ratio, changes in geographic incidence rates, diminishing effect of age at migration and MS risk, and birth month and MS risk.

The aetiology and pathogenesis of MS are still unclear. We can not really state that MS is an auto-immune disease, though many people do. Genes and environmental factors lead to immune damage and CNS degeneration. Environmental manipulation could alter disease risk and/or prognosis (hygiene, infection, vaccines, climate, vitamin D, smoking, diet).

There is a possible unifying environmental hypothesis of smoking, vitamin D deficiency, hygiene, EBV or other infections for initiating MS, SLE, and RA in genetically susceptible individuals.

EBERS, Oxford, UK: Time of exposure of environmental factors in multiple sclerosis

In contrast to what is seen with regard to environmental factors, there is an accumulative effect on MS risk based on the level of shared genes. This is illustrated by the following table:

This raises the question of the inheritance pattern. Is it a pattern that has been closest to ‘dominant', but only for two generations. The brother and sister rate is very close to that for a child or parent (3.5%) which means that there is virtually no dominance variance. Yet 3^rd and (especially) 4^th generation pedigrees with MS are very rare. So on a population level the MS rate has increased, while at a family level (between individual pedigrees) this rate has gone down considerably. What is the explanation for this reciprocal development?

When looking at the cumulative effect of a shared environment, it is impossible to detect any effect of the shared familial or cohabitional microenvironment, with one single exception (which is the DZ sibling risk). So geographical factors (such as climate and diet) that determine huge effects at a broad population level do not discriminate among familial microenvironments.

As mentioned in an earlier presentation, Norway is a discrete exception to the latitude gradient (amount of sunshine and vitamin D). It illustrates that there could be some kind of interaction between environmental factors (such as climate and diet). Over the past decades the observed change in MS sex ratio has occurred almost unnoticed. In fact the sex ratio is actually not far off from being equal, but on top of that gradual changes in the environmental have selectively increased MS risk in females. The effect of migration on sex ratio has been studied among CCPGSMS patients. Yet unpublished data reveal substantial differences in sex ratio's between immigrants that moved to Canada before or after the age of 21. MS sex ratio's (both for <21 and ≥21 years) among immigrants from northern Europe (2.89 and 1.94) are higher than those among immigrants from southern Europe (2.33 and 1.63).

But also the season of birth seems to have an effect on MS risk. In Canada there is an increased MS risk of 10% when being born in May and a decreased MS risk of 10% when being born in November. This effect is reversed in Australia and Argentina (Ponsonby and Luetic, unpublished) and most importantly the unaffected siblings deviate as significantly from population controls as do their MS affected siblings, but only in the opposite direction. This indicates that something very early in life is influencing the MS risk.

Inheritance study strategies

Determining risk variations in half siblings (who have one parent in common, not two) is a modern approach of studying genetic epidemiology in MS. This approach allows us to study all of the following questions (‘one stop shopping'):

- Does it matter which is the common parent (father or mother)?

- Is there a difference in risk between ½ siblings raised together or apart?

- How many genes are involved in determining the risk?

From the CCPGSMS cohort it was found that the ½ sib rate was equal for those raised together or apart. There is a clear parental origin effect, in which the mother is much more likely to be the common parent. What came as a shock was that the full sib rate in ½ sib families was an additional 3.5% (confirmed by four different and completely independent studies in Canada). So the expectation that the ½ sib rate would plummet didn't come true. Instead the ½ sib rate (1.9%) was actually more than half that of the full sib rate, which implies that the effect of non-MHC genes on inheritance in families is very small.

Another way of studying inheritance is by looking at family relations between aunts, uncles, nephews and nieces (AUNN-pairs). From the 27,900 MS patients in the CCPGSMS study, 1,116 AUNN pairs were identified. It was found that if the niece was related to the affected aunt or uncle through the unaffected mother, then this increased the MS risk (Herrera, 2008).

There may be at least 3 or 4 environmental effects on MS risk. The month of birth effect is not gender specific so it is different from the sex ratio determinant. There also is a distinct maternal effect, though it is not related to the month of birth and so must be prenatal. The effect of migration (which is country of origin specific in migrants) could be genetic but there is no evidence and may therefore well be environmental. The interaction of environment with genetics is haplotype-specific which is probably not true of the preceding factors.

The maternal effects (or even before) are characterized by maternal alleles (Chao). The month of birth effect is maternal or neonatal. The migration effect is clearly neither maternal or neonatal and is age-limited (<20 years). The triggering effect could be at any age but is profiled by intrinsic age-related risk.

HAAS, Berlin, Germany: Women's lifestyle and multiple sclerosis

Data related to lifestyle and MS have been collected by the German MS Registry and the Berlin Quality of Life study. In the German MS Registry data were reported by neurologists from 102 centres across Germany from 2005–2008. In total 15,578 data sets were evaluated. In the Berlin Quality of Life study questionnaires were answered by MS afflicted members of the MS Society Berlin in 1995. In total 673 data sets were evaluated.

Baseline data from both study populations showed a clear difference in EDSS and use of immunological treatment. In the German MS Registry 63% of the study population had an EDSS < 4, and 78% used immunological treatment. In the Berlin QL study these percentages were 13% and 18.4% respectively. Mean age (44 vs. 47.4), mean duration of the disease (12.3 vs. 15.4) and amount of women (71.4% vs. 76%) were comparable.

From these studies no gender related differences in lifestyle were observed regarding partnership, access to the health system (frequency of contacts with physicians, inpatient frequency, frequency of rehabilitation), membership of MS Society, participation in MS groups, and staying in nursing homes. 

Gender related differences in lifestyle were observed regarding employment (more men in full time jobs, more women part time employed or house wife), smoking (M>F), alcohol and drug abuse (M>F), use of complementary medication for MS (F>M), sexual and urinary disturbances (M>F, both reported and treated), need for professional care (F>M), and rating of handicap (lower in women).

From these observations the following conclusions were drawn:

- lifestyle with MS seems to be more influenced by the severity of the disease than by gender.

- some differences cannot be explained by MS alone and may be similar in the general population.

Next step is to perform a subgroup analysis concerning severity and age.

CABRE, Fort de France, Martinique: Manipulation of environment and multiple sclerosis

The French West Indies include Martinique and Guadeloupe. Their Afro-Caribean population is similar (>90%), and both islands are volcanic. A study on African-European interbreeding from which the French West Indian population was derived, has shown a Caucasian admixture of genes of about 27%.

In the last 15 years (1992 – 2007) the spectrum of demyelinating diseases of CNS in the French West Indies has reversed. MS has emerged in the early 1990's and its incidence has continuously increased. In 1998 MS prevalence was 17.4 per 100,000 inhabitants. Between 1992 and June 2007 the crude MS incidence was 1.3 per 100,000 inhabitants and the adjusted incidence (based upon the world population) was 1.2. During this period 130 MS cases were observed, with mean age of onset 34.2 years (range 13-56) and 82% RRMS, 11% SPMS and 7% PPMS. For both sexes MS incidence rate were highest in age groups 25-34 and 35-44. The MS incidence rate rose 0.74 between 1992-1997 and 2002-2007. Between 2002-2007 the MS incidence rate was 1.67 per 100,000.

During the same period NMO has had a stable incidence (0.2 per 100,000) and still affects almost exclusively women. A first wave of MS incident cases in women was later followed by men. F/M ratio in MS incidence cases according to decade of birth, showed the highest ratio's in the 1950's and 1960's groups (6.5 and 5 respectively). F/M ratio's in the 1970's (2.1) and 1980's (2.8) groups were substantially lower.

Interestingly, MS incidence rates increased more among the Martinique population than the Guadeloupe population. In the period 2002-2007 MS incidence rate in Martinique was 2.3 per 100,000 against 1.05 in Guadeloupe. Over time the societal conditions have changed. Before the 1980's the population was mostly born and raised in the French West Indies. They were usually farmers living in a rural and traditional habitat with poor sanitation. Nowadays, the population is better educated, some have moved to Continental France, they are urban citizens, work in the tertiary economy (having office jobs) and most of them have access to clean water.

Three environmental changes are suspected to be (partly) responsible for the observed changes:

- Migration to mainland of France

- Change in sun exposure behaviour

- Improved sanitation (clearance of helminthiasis)

POHL, Ontario, Canada: EBV and multiple sclerosis

Over one hundred years ago Pierre Marie stated that "the cause of insular sclerosis is intimately connected with infectious diseases”. Emil Pfeiffer gave a clinical description of infectious mononucleosis (Pfeiffer's disease) in 1889. Yet, it was only in 1964 that the Epstein Barr virus (EBV) was discovered by Michael Epstein & Yvonne Barr in cells cultured from a lymphoma specimen sent to them from Uganda by Denis Burkitt. In 1968 EBV was demonstrated as etiological agent of infectious mononucleosis.

EBV is a B-lymphotrophic human DNA herpes virus. It affects more than 90% of individuals world-wide and has a life-long viral persistence. Infection goes via exchange of saliva, mainly in infancy (via mother or other kids, e.g. shared toys) and is mostly asymptomatic. In puberty it is often symptomatic, therefore called kissing disease or infectious mononucleosis. The infection and persistence pathways are still not fully understood. EBV somehow manages to pass the epithelium and infect B-cells. These B-cells are then stimulated to divide. In the active phase of the infection as many as 1% of all B-cells are infected by EBV. This then provokes a very strong non-specific T-cell activation of atypical lymphocytes. The virus specific T-cells then manage to kill most of the infected B-cells, but some of the B-cells manage to survive and enter into a persistent phase. From time to time some of these latent B-cells enter a lytic stage and then free virus is released that can infect other individuals.

There are many EBV associated diseases, both malignant and non-malignant. Malignant diseases (with an immunocompetent host) include B-cell malignancies (Burkitt lymphoma, Hodgkin lymphoma), T-cell malignancies (extranodal NK/T cell lymphoma, hemophagocytic syndrome T cell lymphomas) and epithelial cell malignancies (nasopharyngeal carcinoma, lymphoepithelioma-like carcinoma.

Nonmalignant diseases include infectious mononucleosis, oral hairy leukoplakia and perhaps also autoimmune diseases (such as rheumatoid arthritis, SLE, MS).

MS & infectious mononucleosis share a similar prevalence distribution. MS is more common in regions with a higher prevalence of mononucleosis. Regions with low hygienic conditions have a lower prevalence of mononucleosis, because most children are infected during infancy and do not get infectious mononucleosis. According to meta-analysis of all published studies until 2006 a history of mononucleosis provides a 2-fold increase of MS risk (Thacker, 2006). From a Danish follow-up study among 25,234 Danish patients with mononucleosis a 2.3-fold increase in MS incidence appeared (104 patients developed MS). MS risk increases within 5 years of mononucleosis and remains increased for more than 3 decades (Nielsen, 2007). Individuals after infectious mononucleosis have a 20-fold higher MS risk than those who have a EBV negative history.

EBV seroprevalence rates are higher in MS patients than in controls. Virtually all MS patients (99.5%) have an EBV positive background (Ascherio, 2007). Also EBV antibody titres are increased in MS patients. This applies for both anti-VCA (viral capsid antigen) and anti-EBNA-complex (nuclear antigen). Interestingly, an increase in anti-EBNA-1 seems to precede clinical onset of MS by 5 to 20 years.

Additional hints for an association between EBV en MS are that:

- intrathecally produced IgG of MS patients reacts with EBV protein

- EBV specific T-cells are increased in MS

- some EBV specific T-cells cross react with myelin antigens

- CNS B-cells of MS patients harbour EBV

Despite these clear associations between EBV and MS the following questions remain open for further study:

- Is EBV infection a conditio sine qua non for MS? Are there MS patients without an EBV infection?

- Why do only a minority of EBV positive individuals develop MS? Is there perhaps a ‘MS-variant' of EBV? Do genetic susceptibility factors play a role? Are there other pathogenic co-factors (such as smoking, hormones, second infection)? Are there protective factors (such as vitamin D)?

- How & when do EBV infected B cells enter the CNS?

- What causes the CNS damage? Is it bystander damage or a cross-reaction with myelin antigens?

- And then: What was first? Is it EBV that causes MS or is it an immuno-dysregulation leading to MS and an altered EBV immune response?

PERRON, Geneva, Switzerland: The triggering role of herpes viridae in multiple sclerosis

Herpes viruses (HV's) are infectious agents that are most frequently associated with MS. HV's (such as Alpha HV: HSV-1 and VZV / Beta HV: HHV6 / Gamma HV: EBV) are significantly detected in MS versus controls. Direct detection of HV has been reported at certain disease stages, in certain tissues and in certain patients. There are probable co-factors triggering downstream pathogenesis. These associated co-factors could be triggering autoimmune inflammation, or perhaps something else. We now know that HV transactivators trigger the activation specific promotors of the Human Endogenous Retrovirus Type W (HERV-W) multicopy gene family.

In the late 1980's it was suggested that a novel HTLV retroviral member could cause MS (Koprowski, 1985). When a new human retrovirus was isolated in MS (Perron, 1991 / Haahr, 1991), everybody thought that a new HTLV was discovered. Indeed, it was a new MS retrovirus, but not a new member of the HTLV family (Perron, 1997, P.N.A.S.). Instead is was from a novel family of retroviral endogenous elements, HERV-W (discovered from MS retroviral sequences) (Blond, 1999 / Perron, 2000). This introduced a novel context at a time when half of the human DNA was yet unseen. Later, when the human genome was completed it was found that HERV's are 8% of the human genome. This opened up a new domain in human biology: HERV's are both viruses, genes and pathogens altogether. And so we must understand the physiopathology of these transposable elements, their interplay with infectious co-factors and their consistency with diseases such as MS.

The expression and accumulation of HERV-W ENV protein has been detected in MS brain lesions (in plaques) by several independent studies. These studies did not detect HHV6 sequences and protein in the established lesions. This matches with everything that has been said about HV (including EBV), that they act as some kind of trigger early in or just before the onset of MS. In this ‘hit-and-run' theory HV's act as an indirect trigger for MS, by triggering HERV-W, which reaches incredible levels of copies in MS plaques.

There are many publications that have associated different types of HV (HHV6, VZV, HSV-1, EBV) and some bacteria (C. Pneumoniae) with MS disease. But a direct causality between these neurotropic agents and MS has never been proven. It may well be that HERV-W is the missing link that leads to MS. Several studies have shown that these neurotropic agents together with a pathogenic copy of HERV-W could activate neuro-inflammation and eventually MS. MSRV ENV Protein from the HERV-W family has been detected as a neuro-inflammatory immunotoxin (Perron, 2001 / Rolland, 2006). Nowadays HERV-W & MSRV ENV transactivation by HV's is supported by several studies.

So we have HV's that can trigger expression of a pathogenic copy of the HERV-W family when present in susceptible individuals. When expressed the ENV protein has an immunopathogenic effect and also induces anti-myelin autoimmunity. Interestingly, the sequence of these ENV proteins is matched by members of the endogenous families. This means that they have multiple ‘brothers and sisters' inserted in many chromosomes. It appears that the human X-chromosome is an important co-activating or amplifying region that adds to the expression of the ENV protein. Critical copies on the X-chromosome with ENV sequence were nearly identical to that of ‘MSRV' copy. But the question remains whether this also means that gender makes a difference.

It has been established that there are different triggering co-factors. If you neutralize one, another will take its place. Therefore it is better to target the ENV protein that is expressed from a pathogenic copy of HERV-W. This ENV protein triggers the pro-inflammatory cascade by activating the TLR-4 receptor pathway. By blocking this pathway this activation of the innate immunity and autoimmunity cascade can be prevented. The presence of the therapeutic target (ENV protein) has been confirmed in a study among 75 MS patients. GeNeuro is now developing an antibody that neutralizes the ENV protein in order to block the pro-inflammatory cascade and HERV-W accumulation.

ALONSO, Minneapolis, USA: Oral contraceptives and the risk of multiple sclerosis

As other speakers have discussed earlier, there are sex differences in MS incidence and also pregnancy has an influence on MS risk. In some MS animal models the administration of exogenous estrogens has shown to decrease the risk of MS as well as slowing the development of MS. There is a high prevalence of use of oral contraceptives in women during a period in their lives that they are also more susceptible of developing MS. This is why it is interesting to study a possible association between the use of oral contraceptives and the risk of developing MS. The presented evidence comes from the following epidemiological (prospective, cohort) studies:

In this study, published in 1993, 17,032 white women from 17 clinics in Britain had been recruited between 1968-1974. These women received annual follow-up calls until 1991. Information on the use of oral contraceptives and reproductive outcomes were collected at baseline and during follow-up. In total 63 new cases of MS were observed during follow-up. In this study the MS risk of oral contraceptives users was lower than that of non-users, but this difference was not statistically significant.

In this study, published in 1998, 46,000 women were recruited through their family doctors. These women were followed-up from the late 1960's until 1996. Information on the use of oral contraceptives and medical history was updated by their doctors. In total 114 new MS cases were observed in this study. The MS risk of oral contraceptives users was slightly higher than that of non-users, but this difference was not statistically significant.

In a first study 121,700 US nurses, aged 30-55, were recruited in 1976. In a second study started in 1989 116,671 US nurses, aged 25-42, were recruited. Information on the use of oral contraceptives and the development of MS was collected through mailed questionnaires (every two years). 315 MS cases were reported after review of medical records. A statistical significant MS risk increase was found among women using oral contraceptives between 6-8 years. However, in the context of other study findings any clear association between the use of oral contraceptives and MS risk is unlikely.

The GPRD is a computerized database in the UK, which includes clinical information on more than 5 million people. This clinical information is reported by GP's. In this review the diagnosis of MS was made after reviewing the medical records based upon diagnostic criteria (Poser criteria). Also information was collected on oral contraceptives prescriptions from GP's.

A case-control study was conducted, nested in the GPRD, from 1993-2000. Cases were selected which met the Poser criteria and had at least 3 years of medical information before the first symptoms. Controls were matched by age, sex and practice. A total of 106 MS cases were selected (women under 50 years old) who were matched by 1,001 controls (almost 10 controls to each MS case). In this study the oral contraceptive users (both former and current users) showed a considerably lower MS risk than non-users. Also MS risk turned out to be considerably lower in women who had stopped only recently (less than 12 months) with their use of oral contraceptives than those who had already stopped for a longer time (more than 12 months).

Next steps are to perform a replication of the GPRD results in a different population and to find out whether all oral contraceptives have a similar effect on MS risk.

Oral contraceptives do not seem to be associated with the risk of MS in the long term. However, oral contraceptives may be associated with a lower risk of MS in the short term, delaying the onset of the disease. The public health message from these findings is, that decisions about oral contraceptives use should not be based on the potential risk of developing MS.

CERGHET, Detroit, USA: Sexual dimorphism in the white matter of rodents

Sexual dimorphism in CNS is defined as the existence of differences in the anatomical structure and function of male and female tissues. Social biology is currently established in cognition, memory, language and response to stress. Also it is implicated in susceptibility to CNS disease. Sexual dimorphism in CNS was first demonstrated in 1976 by Nottebohm & Arnold in the vocal control area of the songbird.

About sexual dimorphism in the white matter of rodents the following is already known:

- The size of the corpus callosum (CC) in male rodents is much larger (Fitch, 1990,1991);

- There is an increased percentage of myelinated fibers in genu of CC in males rodents (Mack, 1995);

- The number of myelinated axons in the splenium of rat male CC is increased (Juraska and Kim, 1996);

- The increased size and myelination of the CC continues between adulthood and middle-age in rats (Yates and Juraska, 2007);

- White matter volume and volume of myelinated nerve fibers are larger in young male rats (Yang, 2008);

- It is generally viewed that differences between male and female rodents result from action of different sex hormones;

- Receptors for estrogen, progesterone, and androgen were found in oligodendrocytes (Santagati, 1994).

Oligodendrocytes may be the target for steroid hormones and may mediate steroid action on the CNS in the context of steroid metabolism, brain development and sexual dimorphism. Own rodent study has shown that the density of oligodendrocytes is higher in the CC, fornix, and spinal cord of males compared to females. The number of CA-II+ cells and PLP+ cells is larger in males at all ages and in all studied structures. The number of CA-II+ cells and PLP+ cells at a given age are essentially the same.

Myelin messages and myelin proteins are different in males versus females. Oligodendrocyte-specific markers are increased in males compared to females. Expression of MBP is increased in the brain and spinal cord of male rodents. The total MBP and CA-II is increased 20% in brain and 25% in spinal cord in males compared to females. 21.5 kDa MBP isoform is increased 30-35% in females compared to males. CA-II mRNA is increased 3 fold, and PLP mRNA is increased 1.7 fold in males compared to females.

Proliferation of oligodendrocytes and cell death is increased in females compared to males. The number of generated oligodendrocytes (measured by number of BrdU+ cells) is increased in females because the number of newly generated glia is greater in females. Turnover of oligodendrocytes is increased in female rodents.

In the EAE model testosterone was shown to have a protective role. At physiologic levels testosterone provides disease protection in young men (Voskuhl and Palaszynski, 2001). Also testosterone level was shown to be lower in MS women than in controls (Tomassini, 2005). Additional study has shown that castration of mice has a decreasing effect (of 20-25%) on oligodendrocyte density and a two fold increasing effect on glia cell proliferation. Exogenous hormones, certainly androgens, affect glia generation and formation of new oligodendrocytes in the brain. Alterations in primary sex hormones, presumably testosterone, affects

proliferation and cell death of oligodendrocytes.

TIWARI WOODRUFF, Los Angeles, USA : Neuro-protective and anti-inflammatory effects of estrogen receptors ligand treatment in mice

Nowadays, available treatments enhance the quality of life for people living with MS by modifying the disease course, treating exacerbations, managing symptoms, improving function and safety and providing emotional support. None of these treatments directly protect the CNS. They all have an indirect protective effect by decreasing inflammation.

The observed decrease in relapse rate during late 3^rd trimester of pregnancy has several reasons. Decreased cell mediated immunity and increased humoral immunity promote fetal survival and makes it ready for child birth.

During pregnancy there is a remarkable increase of sex hormones, like the estrogens estradiol, estriol and progesterone. Are estrogens responsible for the improvements in MS during pregnancy? Estrogens are involved in much more than reproduction alone. They also play a role in stress, fluid intake, learning and eating. In the CNS they modulate the blood brain barrier, have a selective increase in cerebral blood flow, have multiple effects on glia cells, stimulate dendritic growth and synaptogenesis, influence connectivity and circuitry and possibly also have neuroprotective capacities. Estrogen is also one of the most potent activators of prolactin secretion from the pituitary gland (Freeman, 2000).

To test the potential beneficial effects of estrogens, various EAE models have been used. In EAE there is a relapsing remitting kind (SJL) and a chronic progressive kind (C57BL/6) depending on the strains and kind of peptides used. The pathology is actually very similar to MS. There is white matter inflammation, white matter demyelination and gray matter abnormalities. It turns out that estrogens are immunomodulatory (cytokines, chemokines, regulatory cells, dendritic cells) as well as neuroprotective in EAE. They result in decreased numbers of inflammatory cells in the CNS. So could estrogens be used as a therapy in MS? It turns out that the striking protective effects of estriol and estradiol in EAE raise the possibility that some form of estrogen might be beneficial as a therapy for MS. Indeed a phase I clinical trial of estriol has demonstrated significant benefit (Sicotte-Voskuhl, 2006).

Questions to be raised here are:

- What are the mechanisms?

- Which estrogen receptors (ER) are involved?

- Is there an indirect protection of the target organ via anti-inflammatory effect? So are estogens affecting immune cells by ERα, ERβ, or membrane effects (mER)? or/and

- Is there a direct protection of the target organ via neurons, oligodendrocytes and astrocytes? So is estrogen affecting the CNS directly by ERα, ERβ, or membrane effects (mER)?

Determining which ER mediates the protective effect of an estrogen treatment is of central importance for future development of selective ER modifiers which aim to maximize efficacy and minimize toxicity.

The objective of the presented animal study was to contrast effects of ERα versus ERβ ligand treatment in EAE.

To determine the effect on clinical outcomes, inflammatory and neurodegenerative changes two ER ligands were used. The ERα ligand, called PPT, is 100 fold more selective for ERα. The ERβ ligand, called DPN, is 40 fold more selective for ERβ. ERα ligand treatment was likely to be both anti-inflammatory and neuroprotective.

From a set of experiments the following observations and conclusions were made concerning ERβ ligand treatment:

- ERα ligand therapy reduces EAE scores early and late in the disease, while ERβ ligand therapy reduces EAE scores only late in the disease.

- ERα versus ERβ selective ligand therapy has differential effects on auto-antigen specific immune responses in C57BL/6 mice. ERβ selective ligand therapy does not have an effect on immune response.

- ERα ligand therapy is anti-inflammatory, while ERβ ligand is not.

- ERα and ERβ ligand therapy both reduce demyelination pathology in spinal cords of mice with EAE

- ERα and ERβ ligand therapy both reduce axonal pathology in spinal cords of mice with EAE

- ERα and ERβ ligand therapy both reduced neuronal pathology in spinal cords of mice with EAE

- DPN mediated neuroprotection is dependent upon ERβ

- ERβ ligand therapy results in recovery of motor function late during EAE

Unlike ERα ligand treatment, ERβ ligand treatment is protective within the CNS, independent of anti-inflammatory effects in the peripheral immune system.

Most currently available MS treatments are indirectly anti-inflammatory or neuroprotective. Whereas in the case of estrogens there is a direct neuroprotective effect. ERα ligand treatment is neuroprotective because it is also anti-inflammatory. But ERβ ligand treatment is neuroprotective without being anti-inflammatory.

These data demonstrating that treatment with an ERβ ligand is neuroprotective, are of clinical relevance because breast and uterine endometrial cancer are both mediated through ERα, not ERβ. For neurodegenerative diseases with only a minimal inflammatory component, treatment with an ERβ ligand may suffice. For diseases such as MS with a significant inflammatory component, a standard anti-inflammatory treatment could be used in combination with ERβ ligand treatment. In each of these scenario's, the neuroprotective properties of estrogen treatment could be maintained while avoiding the increased risk of cancer in the breast and uterus.

GREGG, Cambridge, USA: White matter plasticity and enhanced remyelination in maternal CNS of mice

The CNS whiter matter is composed of myelinating oligodendrocytes. The myelin sheath influences neuronal health, conduction velocity, and the synchronicity of spike time arrival in the CNS. One oligodendrocyte can myelinate up to 40 axons and so one oligodendrocytes can play quite a substantial role in a variety of functions.

New oligodendrocytes are generated throughout adulthood. They are generated by oligodendrocyte precursor cells (OPC) that are present and mitotically active throughout the adult CNS. These new oligodendrocytes that are born in the adult brain myelinate axons. The generation of new oligodendrocytes contributes to brain plasticity and myelin regeneration in the adult CNS.

From the field of neurogenesis we have learned some of the physiological mechanisms regulating the birth of neurons. The two principal regions for neurogenesis are the hippocampus and the subventricular zone. We have learned that environmental enrichment, exercise and stress can influence the generation of new cells in the hippocampus. In the subventricular zone the generation of new cells is influenced by estrus, odour exposure, pheromones and pregnancy. The role of white matter in brain function and the physiological changes that lead to the generation of new oligodendrocytes in the adult brain are not entirely clear. This makes it difficult to hypothesize which processes might dramatically influence the generation of white matter in the adult.

MS affects women more often than men. MS female patients experience disease remission during pregnancy.

MS consistently remits during the third trimester of pregnancy, but relapses during the postpartum period. A mouse model of MS (EAE) similarly experiences remission during pregnancy. MRI reveals that pregnancy-induced remission is associated with decreased number and size of active white matter lesions (van Walderveen, 1994). The mechanism of remission is not fully understood, but is thought to be due to depressed T-cell-mediated (Th1) immunity during pregnancy.

We asked ourselves whether pregnancy promotes the generation of new oligodendrocytes and enhanced myelin repair in the maternal CNS. From a series of animal experiments we learned that:

- Pregnancy promotes OPC proliferation in the maternal CNS;

- Pregnancy results in an increase in the generation of new oligodendrocytes throughout the maternal CNS;

- Newly generated oligodendrocytes appear to generate myelin in the maternal CNS;

- Pregnancy enhances the generation of new oligodendrocytes in response to demyelination;

- Pregnancy promotes the remyelination of demyelinated axons, which significantly enhances the repair of demyelinated lesions in the CNS;

- Pregnancy-induced OPC proliferation requires prolactin receptor (PRLR) signalling;

- Prolactin promotes OPC proliferation and the generation of new oligodendrocytes in vivo;

- Prolactin treatments promote an increase in the generation of new oligodendrocytes following demyelinating lesions;

- Prolactin treatments promote the repair of myelin damage in the adult CNS.

Conclusion

The collected data suggest that during pregnancy a surge in prolactin promotes the proliferation of OPC's and the generation of new oligodendrocytes throughout the maternal CNS. Newly generated oligodendrocytes appear to mature and generate myelin during late pregnancy and into the postpartum period. The generation of new oligodendrocytes during pregnancy is associated with an enhanced ability to repair demyelination, which is mimicked by prolactin treatments in virgin females.

The function of the described phenomena is not clear. Does the increased generation of oligodendrocytes during pregnancy function to accelerate the maturation of the female CNS? Is this perhaps a form of plasticity that is preparatory for maternal behaviour? And in context of MS and new treatment opportunities, do high levels of estrogen during pregnancy open a window for prolactin-mediated regeneration in women suffering from MS by suppressing cell-mediated immunity (Th1)?

KUHLMANN, Münster, Germany: Gender differences in de- and remyelination in multiple sclerosis

Axonal damage

Axonal loss is a second mechanism leading to axonal damage. Previous study has shown that there can be significant axonal loss (up to 60%) in chronic MS lesions (Mews, 1998 / Lovas, 2000). No significant gender related differences could be determined in axonal density in early MS lesions. However, in one previous study a significantly lower nerve fiber density was found in the (small axonal fibres at C3 and T2 level of the) spinal cord of male versus female MS patients (Ganter, 1999).

Cortical demyelination

Cortical demyelination is extensive in patients with progressive MS (PPMS, SPMS) and not in relapsing, remitting MS (RRMS). Cortical demyelination can occur in the near absence of focal white matter lesions. Cortical demyelination can be differentiated into three morphologically different types:

- Leucocortical lesions (Type 1), which involve both the grey and immediately adjacent white matter;

- Intracortical lesions (Type 2), which lie purely within the cortex;

- Subpial lesions (Type 3), which can be very extensive and run parallel to the pial surface.

Most frequent types in MS are subpial lesions (44%), followed by leucocortical (38%) and intracortical lesions (18%). Analysis of the number and distribution of types of cortical lesions have showed no gender related differences (Wegner, 2006).

Remyelination

From literature of animal studies we learn that there is a higher density of oligodendrocytes in male than in female mice (Cerghet, 2006). Also the number of dying and proliferating oligodendrocytes in female is higher than in male mice (Cerghet, 2006). There is a higher extent of remyelination in older females than males in toxic demyelination (Sim, 2002). In vitro estradiol protects against peroxynitrite, TNFa and oxygen-induced oligodendroglial death (Marin-Hustegge, 2004 / Takao, 2004 / Canarella, 2004 / Gerstner, 2007).

In order to find out whether the difference in density of oligodendrocytes between male and female mice is similar in humans, a study was performed using a combination of transcription factor Olig2 (expressed by OPC's and mature oligodendrocytes) and NogoA, a marker for mature oligodendrocytes only. Olig2 is an important denominator for the differentiation of neural progenitor cells into the oligodendroglial lineage. OPC's have high levels of Olig2, while mature oligodendrocytes have low levels of Olig2. Analysis of the levels of Olig2 and NogoA in the corpus callosum of humans aged 21-30 and 71-80 provided no indication of gender differences in numbers of OPC's and mature oligodendrocytes. In a group of MS patients OPC levels were higher in the PPWM than in the WM of normal controls, but no gender differences in numbers of OPC's were detected between female and male MS patients. The same outcome was found with regard to gender differences in differentiating and mature oligodendrocytes.

When analysing the rate of remyelination both in early and chronic MS patients, there was a tendency towards more extensive remyelination in females (in both groups). However, none of these finding reached the threshold of statistical significance.

In conclusion, major gender differences do not appear to be present in MS pathology.

COMI, Milan, Italy: Are there gender effects in disease modifying treatments?

A large number of previously performed clinical trials with disease modifying treatments were reviewed for information about possible gender effects. Findings from this review are presented here today.

Gender differences in clinical and MRI findings

With regard to clinical and MRI findings the following gender effects are found in MS:

- In RRMS men have a tendency for more residual deficits after a relapse and a faster disability progression from EDSS 0 to EDSS 3. The speed of disability progression between EDSS 3 to EDSS 6 is similar for both sexes.

- RRMS males have a tendency to a higher T1/T2 ratio in brain MRI

- Women have higher Gd+ lesion numbers in RRMS

- In PPMS no gender related differences are observed

Gender differences in patient inclusion

Female predominance in MS (approximately 3:1) is reflected by the percentage of females included in most of the past clinical trials.

- In CIS clinical trials the female percentage averaged 70.5%

- In pivotal RRMS clinical trials the female percentage averaged 71.6%

- In recent phase II RRMS clinical trials (in recent 3 years) the female percentage averaged 69.4%

- In SPMS clinical trials the female percentage averaged 62.1% (with exception of the mitoxantrone trial, in which also patients with relapsing progressive MS were included)

- In PPMS clinical trials the female percentage averaged 55.2%

These last two observations suggest that the male proportion may be higher in the progressive phase of MS than in the relapsing remitting phase.

Gender differences in CIS treatment

Evaluation of gender differences in response to CIS treatment provides the following results:

- Males respond better in ETOMS

- Females respond better in PRECISE

- No differences are observed between males and females in CHAMPS and BENEFIT

These observations suggest that males respond better than females at very low doses of interferons, while females benefit more than males from the highest doses of interferons. These observations also learn us that various treatments may have different effects in females and males.

Gender differences in RRMS treatment

IN RRMS treatment there are very few data available to make a comparison between females and males. The available data show no gender differences. With respect to natalizumab (AFFIRM) it is clear that males and females are responding similarly to this treatment. But there is a trend towards a better therapeutic response by females as measured by annualized relapse rate and by disease progression after 2 years (not statistically significant).

In the recently completed FORTE study no differences in relapse rates were observed between genders in low (20 mg) or high (40 mg) dosages of GA/9016. However, the MRI-results show a more than 50% decrease in Gd+ lesions and new T2 lesion load in males compared to females. These results are highly significant and are very important in determining treatment strategy.

In the CORAL study (with oral GA/7023 in RRMS patients) a positive, dose-dependent trend in treatment effect is visible only in males (measured by number of relapses, T1 Gd+ lesions and new T2 lesions). Although this positive trend in males is not statistically significant, it is absent in females.

Gender differences in SPMS treatment

The SPECTRIMS study has shown a better time to progression in male SPMS patients treated with placebo (hazard ratio=0.64), which was statistically significant. In the treatment arms females responded better than males as measured by time to confirmed EDSS progression, T2 activity measures and % change (on MRI) in burden of disease. In all these measures the ‘treatment x sex' interaction proved statistically significant.

Conclusion

From this review we learn that the proportion of females included in SPMS trials is lower in comparison to RRMS clinical trials. There is no clear gender effect on the response to treatment in CIS. Also there are no gender effects on the response (in clinical markers) to GA and interferon's. In GA treated patients a higher MRI activity was found in females and a trend towards a better MRI response in males. Female MS patients respond better to treatment with either natalizumab (AFFIRM) or INF (SPECTRIMS).

From this review we can summarize that different gender related effects were observed according to types of treatment. The main message from this review is that in the future we have to give more attention to gender related differences in treatment outcomes when determining optimal treatment strategy.

WOLINSKY, Houston, USA: Glatiramer acetate treatment in primary progressive multiple sclerosis. Why males appear to respond favourably

In clinical trials we forcefully introduce severe selection biases (called entry criteria) that are avoided in the best natural history cohorts (complete case ascertainment). In clinical trials we randomize in order to try to understand outcomes. That procedure is often less applicable to natural history cohorts.

The publication of the results of the PROMiSe trial, some two years ago (Wolinsky, 2007), prompted a series of questions. Before addressing these questions it is important to summarize some of the backgrounds of the PROMiSe trial, a multinational, multicenter, double-blind, placebo-controlled trial determining the therapeutic effects of GA in PPMS patients.

The PROMiSe trial enrolled nearly as many males (48.8%) as females. The population was mostly Caucasian (89.8%), their average age was 50 years, time from 1^st symptoms was about 11 years and time from diagnosis was 5 years overall. The trial involved 943 participants from the US, Canada, UK and France.

Another trial conducted more recently, called the Olympus trial, included 439 patients, of which 147 were randomized on placebo. The patient population was quite similar to that of the PROMiSe trial.

In the PROMiSe trial the average EDSS score of the patients at point of entry was 4.9, which was also quite similar to that of the Olympus trial.

The PROMiSe trial started in July 1999 and it took 16 months to enrol 943 subject in the trial that were randomized 2:1 on GA or placebo. The originally planned duration of the trial was three years. In case of overwhelming effect a first interim analysis of blinded data occurred in September 2001, when 722 subjects had been on the trial for more than 1 year. There was no particular reason to stop the trial at that particular point in time. Seeing that the progression rates of the trial patients were slower than projected at the beginning of the trial, estimated, the study was reformatted so that all trial patients would be followed-up until the last patient completed the study. A second interim analysis occurred in November 2002 when a similar amount of patients (757) had been on the drug for at least two years. The dataset monitoring committee felt that a utility analysis was called for and decided that the study results would never come out positive and therefore it was perhaps inappropriate to continue patients on treatment. Patients that had discontinued the use of the drug were asked to be followed-up until they had completed the original 3 year period of follow-up. So at least something could be learned from this important dataset. All the data form this intention to treat study were presented during an investigators meeting on October 8^th, 2003. During that meeting one of the investigators suggested to look at possible gender related differences.

In the data presented 90.8% of all patients had contributed to the data for at least 1 year, 82.5% had been contributing to the data for at least 2 years, but only one third (33.2%) had been contributing to the data for the originally planned duration of the study (3 years). This is important to bear in mind when looking at the results.

No statistical significant difference in disease progression (primary endpoint) was found after 36 months between the GA and placebo groups. In a post-hoc sensitivity analysis on gender it was noticed that male patients on placebo progressed much more rapidly than females either on placebo or active drug as well as males on active drug. There was about a 30% overall reduction in the male group on GA, which was statistically significant.

So from these observed gender differences, a number of questions were raised that can be answered as follows based upon now available data (published and unpublished):

1. Is this an issue of treatment response or of event frequency?

It is likely related to slower progression rates in the females complicating the issue of any clinically measurable effect in this study, but this can only be addressed by another trial designed specifically with this as the central hypothesis.

2. Could the spectroscopy (MRS) data, now published in Multiple Sclerosis, give an answer here?

Substudy cohorts were too small to state anything of importance.

3. Can MRI data can help provide an answer?

The biological effect of GA in this study was not in any way gender specific.

4. Is this phenomenon, as found in PPMS, also present in RRMS?

Meta-analysis (putting together the Bornstein, Johnson and Comi trials) has shown that for the subjects that contributed to the ‘pivotal' core of GA trials, an effect of gender on clinical outcomes was present, but the effect of therapy was not gender dependant.

5. Is this phenomenon specific for GA or also be found in other treatments?

This is difficult to understand without access to primary data, but should always be considered.

VOSKUHL, Los Angeles, USA : Estriol treatment in multiple sclerosis

(presented by Dr. S. Gold)

Rationale for estrogen treatment in MS

MS is considered an autoimmune disease with both inflammatory and neurodegenerative components. The early phase of the disease is predominantly associated with inflammation, focal lesions on MRI and clinical relapses. There is increasing evidence that the neurodegenerative component starts very early in the disease and may at least in part be independent of inflammation. And so an ideal MS treatment should target both components of the disease. However, the main treatments available all have been designed to target inflammation and provide indirect neuroprotection by decreasing inflammation. Nowadays, no treatment is available that is directly neuroprotective. And so the disability process continues to progress despite the effect on inflammation in treated patients.

Based upon currently available information estriol may have anti-inflammatory and neuroprotective properties. And there are additional benefits of estriol over developed MS treatments. Estriol can be administered early as a pill (instead of injection), it is relatively inexpensive and based upon decades of postmarketing experience it is considered to be relatively safe.

There are two studies that have independently suggested that sex hormones may play a role in MS susceptibility and disease progression. The first one suggests that there may be a neuroprotective effect of either male sex hormones (testosterone) or chromosomes. The neuroprotective role of testosterone has been studied in EAE animal models as well as clinical MS patients. There have also been studies on higher susceptibility to auto-immune diseases linked to the X chromosome. Secondly, there has been the clinical observation that MS activity decreases during late pregnancy. This seems to suggest that pregnancy-related factors (such as estriol) may be protective. This decrease in disease activity during pregnancy is not only visible in MS, but also in a number of other cell mediated autoimmune diseases, such as RA, uveitis, thyroiditis and psoriasis. This is due to decreased cell mediated immunity and increased humoral immunity, which in a biological sense are aimed at promoting fetal survival by protecting against an attack.

During late pregnancy (especially in the 3^rd trimester) there is a significant decrease in relapse rate (Confavreux, 1998). This coincides with an increase of several factors including progesterone, estradiol, estriol, cortisol and Vitamin D. From this observation it was difficult to predict which of these factors are responsible for the drop in relapse rate. And so the therapeutic potential of each of these factors (including combinations) was tested in animal models. From these tests estriol and estradiol came out positive in a number of EAE animal models, while progesterone proved to have no effect on EAE.

Clinical trials

When taking these observations to the level of clinical trials, one has to consider efficacy, safety and dose.

Considering safety there are two major estrogen receptors: ERα and ERβ. Erα has an inflammatory and neuroprotective effect, while ERβ only has a neuroprotective effect. Whereas estradiol is much more potent (both towards ERα and ERβ) than estriol, estriol is relatively selective to ERβ and therefore much safer in view of avoiding negative side-effects (like breast cancer and uterine cancer). This makes estriol more suitable to take to the clinical trial.

With respect to dose a previously performed animal study (Kim, Voskuhl, 1999) had already shown that estriol at a low, physiological level during pregnancy is as protective against EAE as higher estriol doses.

Based upon these findings a pilot study was performed. This was a study with a cross-over design among 10 MS patients (6 RRMS, 4 SPMS). Patients were followed-up during 24 months. The first 6 months of pre-treatment were followed by 6 months of treatment, then 6 months of post-treatment and finally 6 months of re-treatment. During treatment periods patients received oral 8 mg estriol tablets each day. There were monthly MRI (Gd, T2) measurements and clinical visits every three months (with assessment of EDSS, 9 Hole Peg Test and blood samples). There were tests for delayed type hypersensitivity (Tetanus, Candida) and PASAT scores were measured.

Some of the findings from this pilot study are:

- The administered estriol was taken up into the blood circulation. Blood estriol levels increased during treatment periods and were low during the non-treatment periods. During treatment periods blood estriol reached levels comparable with estriol levels measured in late second trimester in healthy subjects (Sicotte, Voskuhl, 2002).

- In MRI a significant decrease in volumes and numbers of GD enhancing lesions was observed in RRMS patients during treatment periods. These volumes and numbers increased during the post-treatment period. This effect occurred in RRMS patients and was not observed in SPMS patients (Sicotte, Voskuhl, 2002).

- During treatment there was a decrease in delayed type hypersensitivity, especially for Tetanus and Candida. This is an in vivo function for cell-mediated immune memory response.

- Immune cell composition during treatment showed a significant decrease in CD4+ and CD8 + T-cells and an increase in B-cells. During treatment there was an increase in IL-5 and IL-10 and a decrease in TNFa, pointing at a Th1 to Th2 shift. These observed changes in IL-5, IL-10 and TNFa were associated statistically significant with MRI changes (Soldan, Voskuhl, 2003).

- PASAT scores improved during treatment periods only in RRMS patients and not in SPMS patients (Sicotte, Voskuhl, 2002).

Results from animal studies and this pilot study have confirmed that estriol treatment indirectly protects target organs by decreased cellular immunity and decreased CNS infiltration. With respect to direct protection of estriol on neurons, oligodendrocytes, astrocytes and microglia, animal studies have shown that during EAE estrogens can enhance a number of mechanisms that are considered neuroprotective. Estrogens reduce microglial activation, decrease neuronal cell death, decrease oligodendrocyte cell death, induce neurotrophic factors and increase dendritic spines and synapses. Also from a clinical view there are at least some indications that estrogens can be neuroprotective. These indications are derived from studies on normal ageing and several neurodegenerative diseases (such as Alzheimer's disease and cognition in stroke). From these studies it appears that the ability of estrogens to enhance CNS functions is largely dependent on age and disease stage. The ‘healthy cell hypothesis' suggests that estrogen treatment needs to be started early on when there is no widespread neurodegeneration.

So from these findings it has been shown so far that estriol has a strong anti-inflammatory effect and may have additional direct neuroprotective effects as well.

A phase II double-blind, placebo-controlled multicenter study estriol trial has been started enrolling 150 RRMS patients. This study is funded by the National MS Society and the NIH. It is a 24 months study in which patients receive Copaxone along with estriol 8 mg or placebo. Included are females aged 18-50 years old with a diagnosis of RRMS, who have no other significant diseases, do not take birth control pills and do not smoke. Primary outcome measure are the number of relapses in the 2 years study period. Secondary outcome measures (referring to neuroprotection) are cognition and atrophy visible on MRI.

MYHR, Bergen, Norway: Vitamin D treatment in multiple sclerosis

Vitamin D₃ is synthesized by our skin (through UVB radiation in sunlight) and may also come from dietary intake, especially fatty fish and fortified food. Vitamin D₃ is hydroxylated in the liver to become 25-hydroxy-vitamin D₃ (Calcidiol) and is then turned into 1,25-dihydroxy-vitamin D₃ (Calcitriol) in the kidney. This is the active metabolite fraction in the body. This active metabolite is responsible for the Calcium-homeostasis and bone transfer and also has a potent immunomodulatory effect by reduced antigen presentation, shift from Th1 cytokines to Th2 cytokines and modulation of antibody production.

Vitamin D in MS

Vitamin D is a risk factor for MS. MS prevalence figures in France (Vukusic, 2007) show a geographical distribution which matches the amount of sunlight (UV radiation) in these areas. MS prevalence figures are lowest in areas with high amounts of sunlight. A case control study has shown that outdoor activities during summertime is correlated with a lower risk of developing MS (Kampman, 2007). Two longitudinal ‘prospective' studies from the US have looked at the influence of diet supplementation (in a nurses cohort) and circulating levels of 25-hydroxy-vitamin D₃ (in a military personnel cohort) on MS risk. Intake of vitamin D from supplements was shown to be inversely associated with MS risk. The relative risk of women with vitamin D intake of at least 400 IU per day was 40% lower to those who had no supplemental vitamin D intake. (Munger, 2004). Those with high levels of circulating levels (above 100 nmol/l) of 25-hydroxyvitamin D₃ had a reduced risk of developing MS. This correlation was found only in whites (Munger, 2006).

Some studies indicate that vitamin D levels in serum samples are lower during MS relapses than during stable phases of the disease (Soilu-Hanninen, 2005). Early MS relapse and disability progression seem to be related to lower levels of vitamin D (Smolders, 2008). Another study has found a correlation between Vitamin D deficiency and higher EDSS levels (Van der Mei, 2007). However, it may well be that patients with higher EDSS levels get lower sunlight exposure and as a result show more vitamin D deficiency. Evidence from animal models show that 1,25-dihydroxy-vitamin D₃ prevents development of EAE (Cantorna, 1996).

Gender issues

It is generally said that females have lower levels of 25-hydroxyvitamin D₃. Females need a higher dose of vitamin D than males to achieve the same 25-hydroxy-vitamin D₃ serum levels in women (Hatcock, 2007). A Dutch study has shown that only in women MS risk is influenced by 25-hydroxy-vitamin D₃ levels (Kragt, 2008).

There is also an interest in the interaction between vitamin D levels and smoking, especially in females. Female smokers have a lower dietary intake of vitamin D (Morabia, 2000), and they also have lower 25-hydroxy-vitamin D₃ levels (Supervia, 2006).

Then there is the issue of pregnancy and breast feeding. During pregnancy there is an increased level of 25-hydroxy-vitamin D₃ (Verhaeghe, 1992). These levels are increased especially during the last trimester, which contributes to the reduced disease activity found during pregnancy. During pregnancy and breast feeding there is an increased need for sufficient vitamin D levels for the development of the fetus, for the newborn and for the mother. This is interesting in relation to findings that MS risk is associated to season of birth (Willer, 2005). This suggests a relation between MS risk and lower vitamin D levels in pregnant women due to lack of sunshine exposure during the winter period. A number of animal studies have showed inhibiting effects of 1,25-dihydroxy-vitamin D₃ on EAE, especially in female mice (Spach, 2005 / Cantorna 1999).

Treatment in MS

In a type 1 diabetes study the effect of daily vitamin D supplements intake (2,000 IU/day) during the first year after birth has been evaluated. This vitamin D use proved to reduce the risk rate for type 1 diabetes. This was in some way confirmed by the finding that children suspected of developing rickets early in life had a higher risk of developing type 1 diabetes (Hyppönen, 2001).

As mentioned earlier, we learned from a nurse cohort study that daily use of at least 400 IU vitamin D supplements may reduce MS Risk (Munger, 2004). Another study showed that consumption of cod-liver oil (at a recommended dose of 400 IU vitamin D per day) by people with low sunlight exposure reduces MS risk (Kampman, 2006). Outcomes from a Canadian dose-finding safety study, in which MS patients used an average of 14,000 IU vitamin D per day during 1 year, were that vitamin D use is safe and that it is correlated with a trend towards reduced clinical and MRI disease activity (Kimball, 2007 / Burton, 2008). There is also a number of smaller studies that have indicated safety and/or stabilizing of the disease (Goldberg, 1986 / Nordvik, 2000 / Achiron, 2003 / Wingerchuk, 2005) when administering either vitamin D supplements or 1,25-dihydroxy-vitamin D₃. Animal studies have indicated that vitamin D may protect against EAE, stop (in the early phase of EAE) or reduce disease activity (Cantorna, 1996).

Future perspectives

With respect to dosage of vitamin D, health authorities in the UK and US have recommended a minimal dose of 200-400 IU per day and a maximal dose of 1,000-2,000 IU per day irrespective of age, gender or weight. The Canadian MS dose-finding safety study has proven that daily use of 14,000 IU is safe (Burton, 2008). There is a publication in which a general safety recommendation of 10,000 IU per day is given (Hatcock, 2007).

With respect to serum concentration, there is general consensus that a concentration below 40-50 nmol/l of 25-hydroxy-vitamin D₃ is considered vitamin D deficiency. In order to maintain bone structure homeostasis (to prevent osteoporosis and bone fractures) a serum concentration level of 70-80 nmol/l of 25-hydroxy-vitamin D₃ is required. This requires a dose of approximately 1,000 IU per day. The optimal serum concentration for immunomodulation lies probably above 100 nmol/l of 25-hydroxy-vitamin D₃. This requires a dose of 2,000 – 4,000 IU per day.

With respect to determining target populations, vitamin D may play a part in protection (from conception, birth until onset and diagnosis) as well as disease modulation (after diagnosis). In protection the focus must be on high risk persons, including siblings and children of MS patients, CIS patients and pregnant MS patients. In disease modulation we must consider whether to focus only on RRMS cases or also on PPMS cases.

In the future we need to gather more knowledge about the mechanisms of action of vitamin D. Also we need more evidence of efficacy and safety of the treatment. This evidence must come from randomised, placebo-controlled clinical trials, in which use of vitamin D is analysed as add-on treatment.

TROJANO, Bari, Italy : Post-marketing of disease modifying drugs. Gender questions

The data presented today about the post-marketing results of gender related effects of disease modifying drugs come from the Italian MS database network in cooperation with the main MS centres in Italy.

Post-marketing is a research tool to complement information provided by RCT's. It is the only way to evaluate long term effectiveness, identify treatment response predictors and long term adverse events.

However, there are also some limits to post-marketing observational studies. Due to lack of randomization

confronted groups might have large differences in their observed (and unobserved) covariates. These covariate differences may lead to biased estimates of treatment effects. And so appropriate statistical methods are needed to adjust the comparisons.

Methods of adjustment used in post marketing observational studies to estimate treatment effects include:

- Matching: to select controls who are ‘matched' with the treated subjects on background covariates that need to be adjusted for;

- Stratification: to group subjects into strata determined by observed background characteristics;

- Regression adjustment: on background covariates.

These methods may not be practical if the adjustment involves a large number of covariates or strata. The Propensity Score (PS) is the most common device currently used to reduce bias in treatment comparisons in observational studies. The PS summarizes all the information from the covariates in one single number (so called balancing score). The PS adjusts for the probability of being assigned to the treatment group rather than the control group given the patient's covariates. This method was used in a publication of an observational study about the impact of INFb treatment on disease progression in RRMS patients (Trojano, 2007)

Two gender questions regarding disease modifying drugs

A first gender related question asked is, whether there is a gender effect in IFNb treatment in observational studies. To answer this question data were analysed from a large sample of 2,570 IFNb treated RRMS patients (1,774 females, 796 males), who were followed-up for 7 years. The statistical analysis included multivariate Cox proportional hazard models in order to test the impact of gender after adjustment for baseline variables. Then as a sensitivity analysis the propensity-matched analysis was used to test the robustness of the results. Finally the recursive partitioning and amalgamation (RECPAM) Cox tree-growing algorithm model was used to identify subgroups of patients in which gender effect might be more significant. This kind of analysis provides the identification of different groups and the evaluation of the rate of outcome differences across these groups. At each partitioning step one covariate is chosen in this method (sex, disease course, age at first IFNb prescription, relapse rate, disease duration and EDSS) and its best binary split to maximize the difference in the outcome of interest. The statistical analysis was concentrated on three common outcomes: time to confirmed EDSS 4, time to 1 point EDSS progression and time to 1^st relapse.

This exploratory post-marketing analysis suggests that female RRMS patients show a higher risk in time to 1^st relapse than males patients (the difference is more significant in the subgroup of patients with a lower pre-treatment disease activity). Females patients show a trend towards a lower risk for disability progression than male patients (this difference is significant in the subgroup with younger age (below 32 years) at the start of treatment).

A second gender related question is whether in utero exposure to disease modifying drugs is a risk factor for pregnancy outcome in MS. Data on 432 pregnancies (364 females, 68 males) were collected from small sized samples coming from several Italian collaborative studies. These data included 82 drug exposed patients, 97 patients who had been exposed to drug treatment longer than 1 month prior to conception and 160 patients who had never received drug treatment. All these patients were given a structured interview by which detailed information was gathered about the pregnancy course and its outcomes, as well as possible confounders.

Multivariate logistic and linear models showed no association between IFNb exposure and spontaneous abortion.

IFNb exposure was found to be associated with a higher frequency of preterm delivery and with a lower birth weight of the child. These findings were confirmed by the Propensity Score adjusted models. Considering the pregnancy and fetal outcomes in the 63 males, no significant differences were observed between 31 drug exposed and 32 never exposed.

Conclusion

It is time to use observational data to estimate treatment effectiveness in MS (Trojano, Neurology 2007). RCT's and observational studies do not exclude but rather complement each other. Observational studies allow assessment of the long term beneficial effects of drugs already proven effective in short term RCT's. In addition effectiveness data (i.e. predictors of treatments response, adverse events) derived from observational studies may be used as hypothesis generating to be tested later in a RCT.

CONFAVREUX, Lyon, France: POPART'MUS trial, a child of PRIMS

PRIMS study

The Pregancy and Multiple Sclerosis (PRIMS) study has reported a marked reduction in relapse rate during the third trimester of pregnancy, followed by a marked increase in the first trimester after delivery. From the second trimester onwards the annualized relapse rate does not differ significantly from the relapse rate recorded in the pre-pregnancy year (Vukusic, 2004). The mechanism behind this is a change in immunological status. During pregnancy there is a balance shift from pro-inflammatory Th1 cytokines towards anti-inflammatory Th2 cytokines. From an immunological perspective this can be regarded as a ‘graft tolerance' state. After delivery this balance is reversed dramatically, which can be regarded as a ‘graft rejection' state. This immunological balance shift is accompanied by changes in sexual steroid levels. During pregnancy there is an increase of sexual steroids, followed by a reduction after delivery.

The role of various sexual steroids on prevention and treatment of EAE have been analyzed in animal studies. In vitro studies have showed estradiol, estriol and progesterone to inhibit Th1 responses and enhance Th2 responses. In addition progesterone also plays an important role in remyelination and the protection of myelin sheaths in the peripheral nervous system. The potential role of steroids on remyelination in the CNS remains yet unclear. And so there are various arguments why sexual steroids could be beneficial in EAE.

POPART'MUS study

The Progesterone and Oestradiol in the Post-pARTum for MUltiple Sclerosis (POPART'MUS) study can be viewed as a child of PRIMS. The study objective is to evaluate the efficacy of the combination of a progestative drug

and a low dose of percutaneous estrogen in preventing relapses in the first three months after delivery

in women with MS.

POPART'MUS is a phase 3, European, multicenter, randomized 12-week placebo-controlled double-blind clinical trial with two treatment arms (150 patients in each arm). The study sample size of 300 patients is based on the relapse rates observed during PRIMS (0.7 during pre-pregnancy period and 1.2 in first trimester post-partum).

One group will receive oral nomegestrol acetate (Lutenyl^® 10 mg per day) combined with transdermal estradiol (Dermestril Septem^® 75 μg once a week), while the other group will receive matching placebo treatments.

The treatment begins 24 hours after delivery for Lutenyl^® and 2 weeks later for Dermestril Septem^®. The placebo-controlled double-blind 12-week period is followed by a 12-week open untreated period during which the patients will remain blinded. The total follow-up lasts 24 weeks after delivery.

Patients are included and randomized before week 36 of pregnancy. From delivery they are regularly consulted by telephone and receive neurological exams (at 36 weeks of pregnancy, and 4, 12 and 24 week post partum). A subgroup is also followed-up with MRI and biology sampling.

Inclusion criteria for this study are:

- MS according to MacDonald classification (including CIS fulfilling Barkhof MRI criteria for dissemination in time and in space)

- Relapsing remitting or secondary progressive course of disease

- EDSS lower than or equal to 6.0

- Pregnancy less or equal to 36 weeks amenorrhea at inclusion

There is a long list of exclusion criteria, which include:

- desire for lactation (breast feeding)

- desire for a MS disease modifying treatment in the 24 weeks after delivery

Primary endpoint of this study is the rate of relapses during the first 12 weeks after delivery. In addition there are several secondary endpoints related to clinical and MRI outcome. Patients are recruited through the EDMUS network (250 centers in 30 countries). This trial was started in France (in June, 2005) and later extended to Italy (in March, 2007). Patient enrolment so far has reached a level of 112 patients (status 1 November, 2008).

Although preliminary results from the first 83 patients are available, no relationship to study treatment can be provided yet due to blinding of the study. So far no serious adverse events (SAE) related to study treatment were reported. Two SAE's (pleurisy, lithiasic acute pancreatitis) were not related to study treatment.

LIST OF MEDICAL TERMS

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

Astrocyte Supporting cell in the central nervous system

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

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

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

Epigenetics Heritable traits (over rounds of cell division or transgenerationally) that do not involve changes to the underlying DNA sequence

Glia Supportive tissue of the brain

- astroglia: astrocytes

- oligodendroglia: oligodendrocytes

- microglia: cells of the immune system of the brain

Intrathecal In the corticospinal fluid compartment of the spinal cord

Myeline (sheath) Insulation tissue around neurons to allow conductance

Neural precursor cells Cells that develop into neurons

Oligodendrocyte Cell that myelinates neurons

LIST OF ABBREVEATIONS

ADEM

APP Amyloid Precursor Protein

AUNN Aunts, Uncles, Nieces, Nephews

BBB Blood Brain Barrier

BMI Body Mass Index

CC Corpus Callosum

CCPGMS Canadian Collaborative Project on the Genetic Susceptibility of MS

CDMS Clinically Definite MS

CIS Clinically Isolated Syndrome

CNS Central Nervous System

CSF Cerebrospinal Fluid

DM Diabetes Mellitus

DNA Deoxyribonucleic Acid

DZ Dizygotous

DPN Diarylpropionitrile

E2 Estradiol

EAE Experimental allergic autoimmune encephalomyelitis

EBV Epstein-Barr Virus

EDSS Expanded Disability Status Scale

EDMUS European Database for Multiple Sclerosis

ENV Envelope

ER Estrogen Receptor

F/M Females / Males ratio

GA Glatiramer Acetate

Gd Gadolinium

HERV-W Human Endogenous Retrovirus type W

HLA Human Leukocyte Antigen

HTLV Human T-lymphotropic Virus

ICL Intracortical Inflammatory Lesion

Ig Immunoglobulin

IL Interleukin

IFNb Interferon-beta

IFNg Interferon-gamma

IVIG Intravenous Immunoglobulin

LHON Leber's Hereditary Optic Neuropathy

MBP Myelin Basic Protein

MOG Myelin Oligodendrocyte Glycoprotein

MRI Magnetic Resonance Imaging

MS Multiple Sclerosis

MZ Monozygotous

NAWM Normal Appearing White Matter

NIH National Instute of Health

NMO Neuromyelitis Optica / Devic's Disease

NO Nitric Oxide

OCB Oligoclonal Bands

ON Optic Neuritis

OPC Oligodendrocyte progenitor/precursor cell

PASAT Paced Auditory Serial Addition Test

PEL Progesterone / Estradiol x 0.01 during Luteal phase

PIF Pre-Implantation Factor

PLP Proteolipid Protein

PNS Peripheral Nervous System

PPMS Primary Progressive MS

PPT Propylpyrazole Triol

PPWM Peri-Plaque White Matter

PS Propensity Score

RA Rheumatoid Arthritis

RCT Randomized Clinical Trial

RNA Ribonucleic Acid

RR Relative Risk

RRMS Relapsing Remitting MS

SJL Swiss Jim Lambert

SLE Systemic Lupus Erythematodes

SPMS Secondary progressive MS

TNF Tumour necrosis factor

Treg Regulatory T-cell

UV Ultraviolet

VCA Viral Capsid Antigen

VDRE Vitamin D Response Element

WM White matter

--- END ---