Please can you tell us about the origins of this work?
This work is the product of an extraordinary consortium called the Epilepsy Phenome/Genome project (EPGP). It is a consortium of 27 centres in the U.S. and internationally with several hundred physicians and staff who have worked together to enrol more than 4000 patients with epilepsy and their family members. The consortium has collected very detailed phenotypic information and DNA on these participants. It is really through the work of the EPGP consortium that this study has been made possible.
There is a website you can check for further details: www.epgp.org
Can you please outline the different types of migraine? How does migraine with aura (MA) differ from other types?
Migraine is a specific type of headache. Migraine is a recurrent headache disorder with headaches that generally last between 4 and 72 hours. The typical characteristics of migraine headaches are:
a unilateral location – sometimes they can alternate sides from one attack to the next
pulsating quality i.e. throbbing
mild to severe intensity
often activated by a routine physical activity
may be associated with nausea, photophobia (sensitivity to light) and phonophobia (sensitivity to sound).
Migraine with aura (MA) is a specific sub-type of migraine. Some people have some of their headaches with aura and some without – so people can have both migraines with and without aura. In this study we separated participants with migraine into those who ever had auras from those who never had an aura with their migraine headaches.
The aura consists of reversible focal neurological symptoms. Those symptoms, which can be quite varied, usually develop gradually over 5-20 minutes and they last for less than 60 minutes. Usually the headache will follow the aura symptoms.
The most common aura symptoms are visual symptoms such as scintillating scotoma--bright flickering lights and geometric patterns with impaired vision. Sometimes during an aura patients will be unable to see in a certain part of their visual field. They may also have loss of sensation on one side of the body, tingling sensations, weakness on one side of the body, inability to speak, or vertigo. Those preliminary symptoms that are neurological and transient characterise migraine with aura. It is important to distinguish more dangerous neurological conditions that can cause these symptoms before attributing the cause to migraine.
Please can you give a brief introduction to epilepsy?
Epilepsy is a common neurological disorder. Epilepsy is the occurrence of seizures without acute provocation. So if you for example take some toxic substance, or are withdrawing from alcohol and you have seizures only in that setting – that’s a provoked seizure. A predisposition to acute unprovoked seizures is what constitutes epilepsy. The definition has evolved recently so that some would consider epilepsy to be at least one seizure that occurs without acute provocation. Others still rest on the definition that epilepsy requires at least two or more (i.e. recurrent) seizures.
I have to step back to define what a seizure is. A seizure is the clinical manifestation of an abnormal hypersynchronous discharge of a population of neurons in the brain. When a population of neurons starts firing abnormally all together, this electrical event can produce a manifestation in a person: the clinical features. Clinical features can be subjective – something that the patient notices such as a rising feeling in the stomach, a taste or a smell – and/or objective, such as jerking of the limbs or loss of consciousness.
There has been an increasing effort in the clinical and research communities to understand epilepsy in a broader framework in which not just those electrical events but the entire neurobiological, cognitive, psychological, and other consequences of the condition are taken into account.
Epilepsy is common: 1 in 26 people develop epilepsy at some point in their lives. About a third of people are refractory to treatment with currently available anti-seizure medications. Some people go on to be managed by epilepsy surgery if an identifiable region of the can be located where the seizures originate from and that region can be resected surgically. But it still leaves a lot of people that need to be managed either because they are not eligible for surgery or because current therapies don’t work. We really have a long way to go in the treatment of epilepsy.
About two thirds of epilepsy has no identified cause. Genetic factors may play a critical role particularly in that subset of epilepsy. The profound hope of scientists, caregivers, and families with epilepsy is that genetics will offer a completely novel and wider understanding of the causes and the pathophysiology of epilepsy. Understanding more about the genetics of a disorder can help clarify the underlying biology of that disorder. That understanding can be used to develop targeted new therapies, and allow us to really transform the care of our patients. The results of genetic studies of epilepsy are now beginning to improve epilepsy care. This is very exciting for both researchers and clinicians.
How common is the co-occurrence of migraines and epilepsy?
Migraine is more common than epilepsy. The prevalence of migraine in the US was measured twice over ten years and was found to be stable. It depends on gender considerably: it is about 5.5-6% in men and about 17% in women. Across genders it is about 11-12%. Migraine is approximately 10 times more prevalent than epilepsy in the US population.
A number of studies have demonstrated that migraine and epilepsy co-occur more than would be expected by chance. The prevalence of epilepsy in people with migraine is higher than that in the general population, and migraine headache occur almost twice as often in individuals with epilepsy as in individuals without epilepsy. The question that hasn’t been answered fully is whether genetics plays a role in that co-occurrence in the two disorders, migraine and epilepsy. That’s what we set out to examine in the research study I’m talking about today.
Who are migraines and epilepsy most likely to affect?
Migraines are slightly more common in women. The epilepsy gender differences are quite minimal – it depends on the type of epilepsy however.
Anyone at any age can develop epilepsy but there is a bimodal age distribution in epilepsy. The young and the elderly are more prominently affected. 45,000 children under the age of 15 develop epilepsy each year. The incidence is highest in children under 2 and adults over 65, so it really affects all areas of the spectrum but the very young and the elderly are disproportionally affected.
The increase in late age may be a result of some insults that occur over the lifespan and in later life such as stroke; whereas epilepsy occurring before age 35 is much more likely to have an unknown cause and is more likely to be genetic.
Both migraine and epilepsy can occur in children. Migraine is difficult to diagnose in very young children because it depends so much on the reporting of symptoms by the individual who has the headaches. The questionnaire that we used has been shown to be valid for use only in children over age twelve and in adults. For that reason, we limited our study to looking at migraines in adults and children over twelve.
Certain people are pre-disposed to have epilepsy, such as people with mental retardation, cerebral palsy, and stroke. These are people with epilepsy of known cause. The majority of people with epilepsy have no identified cause for their seizures.
What was previously known about the causes of epilepsy and migraines?
Prior to this study, it was clear that epilepsy itself had both genetic and non-genetic contributions. If you have a scar in your brain, you can develop seizures as a result of it. Neurosurgical procedures, central nervous system infections such as encephalitis, stroke, and tumors can cause epilepsy.
The field of epilepsy genetics has really been transformed in the last few decades with the discovery of a large number of epilepsy genes. However, the majority of common epilepsies are not explained by the genes discovered to date. A lot of the progress in epilepsy genetics has come with the help of careful phenotype definition. Phenotype definition involves the identification of specific clinical features--such as seizure type, specific seizure symptoms, and EEG characteristics—in order to classify patients into groups likely to share susceptibility genes.
Successful gene discovery in migraine has been primarily for a rare form called familial hemiplegic migraine, but not for commonly occurring migraine headaches. Familial hemiplegic migraine affects multiple individuals in a single family, and episodes can be very severe and include fever, seizures, prolonged weakness, and coma. Some people have permanent neurological disability after attacks. This is not the case for migraine headaches in general.
There are several pieces of evidence that support the existence of a shared genetic effect on migraine and epilepsy. First there have been some linkage studies. Linkage studies identify a family with multiple affected people with a certain disorder and determine what part of the genome is travelling with the disease phenotype of interest in the family. This results in identifying a region of the genome that is “linked” to the disorder, and therefore likely to contain a susceptibility gene. Linkage studies of migraine, epilepsy, and the combination of the two has been reported to several regions of the human genome. That is supportive evidence for a shared genetic effect, but not confirmatory evidence that one gene causes the two disorders.
The next level of evidence for a shared genetic contribution comes from the fact that genes that have been implicated in familial hemiplegic migraine also play a role in specific epilepsy syndromes. Mutations in several genes including a calcium channel gene called CACNA1A, a sodium channel gene SCN1a and ATP1A2 have been identified in both familial hemiplegic migraine and specific epilepsy syndromes. These studies address a rare form of migraine and relatively rare forms of epilepsy as well, not more common types of epilepsy and migraine.
There hasn’t been much positive evidence from large scale epidemiologic studies to support a shared genetic susceptibility. In fact, several studies failed to show a shared genetic susceptibility including one by Ruth Ottman working with Richard Lipton in 1996 with the Epilepsy Family Study here at Columbia University (EFSCU). That study did not find evidence for a shared genetic effect on the two disorders. I think we understand one of the reasons for this discrepancy with our results. In the epilepsy Phenome/Genome Project, each included family had at least 2 affected members with epilepsy and many had more than that; whereas in EFSCU there were a number of families with just one affected. Because the strength of the genetic effect of epilepsy operating in the family predicts the occurrence of migraine, in families with fewer affected individuals with epilepsy, it may not be possible to detect increased rates of migraine in family members. This may have made the results more difficult to identify in EFSCU than in EPGP.
Andermann and colleagues in 1987 also did not find an increased prevalence of migraine in relatives of individuals with epilepsy compared to controls. There has been one positive study demonstrating a shared genetic effect on epilepsy and migraine in a population of patients with epilepsy, but this study looked at a specific type of childhood onset epilepsy called benign rolandic epilepsy. This study was published in 2009 by Clarke and colleagues.
At the time we began our study of migraine and epilepsy in EPGP, we felt that there was still a need to ask whether this comorbidity of migraine and epilepsy that has clearly been demonstrated in individuals might really be coming from a genetic cause. The Epilepsy Phenome/Genome project was a unique and powerful population in which to study this question.
How did your study into the shared genetic susceptibility of epilepsy and migraines originate?
There had been genes that had been identified that caused both rare syndromes; there had been linkage studies that implicate both. There were some contradictory results. Many thought this question was not fully answered.
Another major impetus for the study was a larger question, which I think is also one of the major implications of this study. There is increasing recognition of the importance of epilepsy and many of its comorbidities, not just migraine. Those recognized co-occurring conditions include psychiatric comorbidities such as depression, anxiety, and suicidality, as well as migraine. The Institute of Medicine (IOM) report that came out last year dramatically emphasized the role of comorbidity in epilepsy. (www.iom.edu) I think we are really starting to understand that epilepsy is not a disorder that stands alone but that has many other manifestations that come along with it-- both consequences and co-occurring disorders – either because of a shared cause or because one causes the other. There is a lot of complicated work that needs to be done to iron that out.
The study of comorbidity in epilepsy is changing the recognition and treatment of patients who have epilepsy and their families. In EPGP we did not measure psychiatric disease but it is something that we are hoping to do – cognitive, developmental and psychiatric comorbidity. But we did measure migraine in the original study using a validated interview that would allow us to diagnose migraine with or without aura in a valid way.
With migraine, we had a perfect opportunity to choose a specific comorbidity that was already represented in our data set. But it really was part of this larger question that is also one of the National Institute of Health (NIH) benchmarks – one of them is this issue of comorbidity and epilepsy. So the treatment and understanding of epilepsy and its related conditions is a hugely important research and clinical question right now.
What did your research involve?
The Epilepsy Phenome/Genome project is an on-going study – we are actually enrolment of patients now. The goal of the Epilepsy Phenome/Genome project is to recruit and perform very detailed phenotyping on, and collect DNA from a large number of participants with epilepsy. The Epilepsy Phenome/Genome project has 27 centers in the US, Canada, Argentina, Australia and New Zealand.
One arm of this study (there are several different sub-projects) focusses on enrolment of affected pairs of people in a family with epilepsy of unknown cause – those are either sibling pairs or parent-child pairs. Participants were enrolled if they and a sibling, or they and a parent, or they and a child both had epilepsy of unknown cause. They are required to have well-characterised epilepsy – generalised epilepsy or non-acquired focal epilepsy.
Generalised epilepsy consists of seizures that originate on both sides of the brain over a wide area of the brain; focal (or partial) epilepsy consists of seizures that begin in a specific area of one part of the brain.
We used this very valuable cohort of families in which 2 or more individuals had epilepsy – either partial or generalised epilepsy of unknown cause. We had very detailed information collected from telephone interviews and interviews in person about participants’ epilepsy and headaches. In our initial interviews we also asked a question about whether there were additional people in the family who had seizures apart from the two people who were already enrolled. We collected that information in a systematic way – everyone in the study was asked – “Is there anyone else in the family who has seizures, who are they and what is their relationship to you?”
We did this initially because we thought it might be useful in case we wanted to go back to these families later and enrol more affected family members, because very large families with multiple affected individuals can be very powerful in approaching the question of genetic cause. We did not enroll those extra patients for this study but we were able to use the information about the additional affected individuals.
Having information on these other people in the family gave us a great opportunity. It is important to note that we did not interview these additional non-enrolled relatives ourselves – we asked the enrolled participants to tell us whether their family members had ‘seizures’. That question has been shown to be valid in identifying additional people in the family who have seizures. Epilepsy is a diagnosis that is best made by talking to the person or their parent if it is a child. This is why it is key to say ‘seizure disorders’ rather than ‘epilepsy’ in this case.
We looked at whether the number of additional individuals with seizure disorders in the family – beyond the enrolled pair - and how close the additional affected relatives were in relation to the participants, predicted the occurrence of migraines in the enrolled participants. Essentially, we wanted to investigate whether a stronger genetic effect on epilepsy in the family (more affected relatives more closely related to the enrolled participant) predicted the occurrence of migraines in the participants. If the number and closeness of relationship of additional relatives with seizure disorders predicts the occurrence of migraine in the participants, that would provide evidence for a shared genetic contribution to migraine and epilepsy.
We had 730 participants aged 12 or older with either generalised epilepsy or non-acquired focal epilepsy. We excluded anyone who had an unclear type of epilepsy. We also only included people who had complete information on migraine. These 730 people came from 501 families so it was a very large sample of strongly familial epilepsy – two or more individuals with epilepsy in the family. That is what really differentiates us from the prior epidemiologic studies that failed to find any effect.
What did your research find?
Our findings can be found in our paper which was published in Epilepsia: http://onlinelibrary.wiley.com/doi/10.1111/epi.12072/abstract
We were able to divide the migraine into migraine with aura and migraine without aura but we didn’t use the standard definitions. We wanted to make sure that migraine with aura and migraine without aura were separate. As they currently are described in the International Headache Society’s classification they overlap. We wanted to say if you never have migraine with aura you will be classified as migraine without aura only, if you ever have migraine with aura you are going to be classified as migraine with aura. We wanted to make sure that we were distinguishing a group of people who were pre-disposed to have aura and those that were never pre-disposed to have aura.
In individuals with epilepsy who had 2 or more relatives beyond the enrolled pair (in other words, in families with 4 or more individuals with seizure disorders), the prevalence of migraines was three times higher than in individuals with no additional affected relatives (families with only two affected individuals with seizures). This means the stronger the genetic effect on epilepsy, the higher the rates of migraine in the participants. This relationship between seizure disorders and migraine only held when we looked at first degree relatives with seizure disorders—when we included more distant relatives the effect did not persist. This is additional evidence that the strength of the genetic effect on epilepsy predicts the occurrence of migraine in these families.
We found an effect when we looked at any migraine, but when we separated out migraine types we found that it was specific to migraines with aura and that migraine without aura did not increase with the number of first-degree affected individuals. This is very interesting pathophysiologically because a lot of people think that auras are a phenomenon that have some similarity to seizures – they are transient neurological disturbances in the brain. The fact that only migraine with aura held up is very interesting and that relationship between migraine with aura and epilepsy has been demonstrated elsewhere.
There are a lot of hypotheses about why migraine and epilepsy might have shared pathophysiology. Some discussion of the biology that links these disorders can be found in the articles found in the links below:
It was also interesting that when we separated out by type of epilepsy in this study – and remember we only knew that for the pairs not for the additional family members – the trend was not specific to epilepsy type. It was also not specific to whether the two participants enrolled in the trial had the same type of epilepsy.
These were very exciting results because it really is a first demonstration of a shared genetic effect on migraine and epilepsy in a large population of common epilepsy and common migraine. This hasn’t been shown before so we are quite excited about this implication. It is also very rewarding for the people who have spent many years doing a lot of work on the Epilepsy Phenome/Genome project to see the power of the consortium’s effort to answer these sorts of questions.
What impact do you think your research will have?
As I said before, one of the ground-breaking aspects of this result is in this larger context of reconceptualising disease boundaries. A disorder does not stand alone but it can be seen as part of a network of intersecting disorders. In fact there have been intersecting bidirectional relationships identified for epilepsy, migraine, anxiety, depression, suicidality and psychosis.
As we start to understand that some of these disorders are occurring in a network or a cluster rather than standing by themselves, I think it is going to completely transform recognition, prognosis, prevention, and treatment strategies. Understanding the shared pathophysiology can direct new therapy of not only epilepsy but the comorbid disorders that may have a tremendous effect on quality of life. If we fail to treat those accompanying disorders we are doing our patients a disservice.
There has been some very strong evidence to show that quality of life and epilepsy depends tremendously, sometimes even more on the occurrence of the comorbidities than on the seizures themselves. We really need to understand epilepsy in its context. There is a huge move in the last few years to do that and I think this work is part of that larger question.
I think we do hope that when you find a common pathophysiology that links two disorders you unravel new strategies for development of novel treatment. Also both migraine and epilepsy are undertreated in the population so the more we can do to help those patients the better sorts of outcomes we can have for patients, their families and the community.
Where can readers find more information?
Our research paper can be found here: http://onlinelibrary.wiley.com/doi/10.1111/epi.12072/abstract
For further information on the Epilepsy Phenome/Genome project please visit: www.epgp.org
Dr. Andres Kanner is an authority on psychiatric comorbidity in epilepsy and has written a number of excellent papers on the topic. Some can be found at:
In the journal Neurology in 2007 I wrote an editorial that reviewed a linkage study into chromosome 9Q, but it also reviewed some of the genes involved in epilepsy and migraine: http://neurology.org/content/68/23/1969
In Neurology in 2010 I also published an editorial about another paper which may be of interest to people. I wrote this with my colleague Dale Hesdorffer here at Columbia University who has done a lot of work on epilepsy comorbidities. This goes over potential biologic mechanisms that link psychiatric disorders, migraine and epilepsy: http://www.neurology.org/content/74/15/1166.extract
A great website for patients is the Internal Headache Society’s website which defines all the migraine terms and valid criteria for diagnosis: http://www.ihs-headache.org/
The Epilepsy Foundation also has great information for patients: http://www.epilepsyfoundation.org/
Would you like to make any further comments?
One thing I would like to emphasize is that genetics is a window into understanding the pathophysiology of disease or the shared biology of two disorders, so it has much wider implications than just finding a gene for a rare syndrome. I think it really is an increasingly powerful way of exploring the reason why these disorders happen. That can really change our ability to help patients and their families.
About Dr Melodie Winawer
Melodie Winawer is an Assistant Professor of Neurology at Columbia University in New York City, and the Director of Clinical Neuroscience Education at Columbia. She is a neurologist and epilepsy specialist in the Columbia Neurology Division. Her research has focused on epilepsy genetics for the past 15 years. She is a translational scientist with expertise in both human epilepsy genetics and the genetics of mouse seizure susceptibility. Her human work has focused on comorbidities of epilepsy, phenotype definition in genetic studies of epilepsy, and the development of novel methods to determine the genetic contributions to specific epilepsy phenotypes.
She is a co-investigator and founding member of the Epilepsy Phenome/Genome Project (EPGP), a large scale multicenter consortium whose goal is to examine genetic effects on epilepsy. She is also the chair of the Comorbidity Core of the Human Epilepsy Project, a consortium which aims to identify predictors of outcome in epilepsy.
New projects include a study funded by the Cure foundation to look at predictors of outcome in infantile spasms, a frequently devastating epilepsy syndrome that begins in infancy. Dr. Winawer’s translational projects in the laboratory have focused on two areas. She has worked towards the identification of genes contributing to mouse seizure susceptibility with the aim of translating these findings into studies of human epilepsy. She is also involved in the development of novel anti-seizure agents that may have effects on both epilepsy and anxiety disorders.
She obtained her MD from the University of Pennsylvania in 1994 and completed Neurology Residency Training at Columbia University/the New York Neurological Institute in 1998. She obtained a master’s degree in Epidemiology from the Columbia Mailman School of Public Health in 2000, focusing on genetic epidemiology. She completed an Epilepsy and Neuroepidemiology Fellowship at Columbia in 2000.
In addition to her research and clinical experience, she plays a major role in education in the medical school at Columbia University, with particular interest in the integration of clinical neurology and basic neuroscience in medical education. In this capacity, she is the Director of the Neurology section of the second year medical student Neuroscience course at the Columbia College of Physicians and Surgeons.