What is sudden unexpected death in epilepsy (SUDEP)?
Currently, the accepted definition of SUDEP is the sudden unexpected witnessed or unwitnessed, non-traumatic and non-drowning death in people with epilepsy, with or without evidence of a seizure. It doesn’t necessarily have to be in the known context of a seizure, excluding documented status epilepticus, where people have seizure after seizure.
Typically, what happens is that someone who’s known to have epilepsy is found to have died without any obvious explanation as to why. Even a post-mortem examination doesn’t reveal a structural or toxicological cause of death. It may or may not be known whether a seizure had occurred and no other cause of death has been identified.
How many people’s lives are taken by SUDEP each year?
In 2013, about 680 deaths were attributed to SUDEP in the UK. If you look at it at a population level, the studies suggest it probably affects about 1 in 1000 patients with epilepsy each year. It depends on the severity of the epilepsy. For example, if you’re looking at people who’ve got epilepsy that’s resistant to drug treatment and who are also not candidates for surgical treatment, then that can rise to as high as 6 to 7 per 1000 people with epilepsy per year.
Why is SUDEP more common in certain types of epilepsy such as Dravet Syndrome?
That’s a good question. I don’t think we have the answer to that question actually and I guess that is partly why we’re doing this research. We know that there are some associated risk factors such as having poorly controlled epilepsy, especially generalized tonic-clonic (or grand mal, convulsive) seizures.
Other risk factors that have been suggested include being male, having an intellectual disability and having seizures that occur during sleep. We don’t necessarily understand the mechanisms behind these factors.
We also know that there are some epilepsies which have an increased incidence of SUDEP and Dravet syndrome is one of those. We know most people with Dravet syndrome have a mutation in a gene called SCN1A, but we don’t know whether the increased risk of SUDEP is linked to having that gene mutation or whether it is due to the epilepsy typically being severe in that condition. We don’t know which of those might be relevant, if either. That’s why people have been trying to understand what lies beneath the risk factors.
Is a lot of research focussing on this area at present?
That’s right. For example, an important study called the MORTEMUS study looked at people who had been admitted to hospital specifically for the recording of their seizures, typically for pre-surgical evaluation. These are people who are in a hospital bed and they are intensively monitored by video. They have EEG electrodes on and they may have their ECG and breathing measured.
A very small proportion of people who have this investigation unfortunately have a seizure followed by SUDEP and the study was able to look at all a number of measures around the time that SUDEP occurred. One of the key findings was that the pre-terminal events were cardiorespiratory, but how those cardiorespiratory events arise is not clear.
More recent work published suggested that there can be a process of spreading depression in the brain stem in the immediate pre-terminal period, so that would fall in line with there being problems with cardiorespiratory regulation systems. However, it still doesn’t tell us why that happens, so although we’re getting close to establishing the mechanisms , we still don’t understand why it happens or why it happens with one seizure and not another.
Are there other causes of premature death in people with epilepsy?
Not all premature death in epilepsy is due to SUDEP. There are other causes and we know that epilepsy can reduce longevity. Among people in whom seizures have been controlled, there can still be a higher risk of mortality at a given age and other factors are involved.
We know other obvious causes of death such as drowning and burns related to seizures, but there are other reasons and we know, for example, that psychiatric comorbidity also appears to be associated with an increased risk of premature death in epilepsy.
What did your recent research reveal about the role genetics plays in a person’s risk of SUDEP?
Because we know that there are some genetic epilepsies associated with an increased risk of SUDEP, such as Dravet syndrome, and because rarely it’s reported in families, we thought that it was relevant to look at the genetic risk for SUDEP and whether there might be genetic factors that increase the risk of SUDEP.
We thought it was unlikely that there is one cause of SUDEP, whether that’s genetic or otherwise, and that it’s probably a combination of risk factors (which may or may not be genetic), a seizure and circumstances. For example, people are more likely to be found in the prone position with SUDEP.
Rather than focusing on a single gene from the outset, we hypothesized that there may be multiple genetic risk factors which somehow add up to increase the overall risk of SUDEP. We therefore took a genome-wide approach to look at the burden of putatively deleterious variants spread across the genome.
We studied DNA from a number of people who had donated their DNA to our research before they had died, as part of our extensive epilepsy genetics studies, and who had unfortunately succumbed to SUDEP later on. That meant we had the DNA of people who we knew had SUDEP, so we were able to study the DNA for possible genetic risk factors.
We undertook whole exome sequencing of these samples and after a series of robust quality control and filtering steps, we identified variants within the DNA that were rare and which were also, on an in silico basis, likely to be deleterious.
What we were able to show was that there was an overrepresentation of such deleterious variants in people who had had SUDEP. We also found that the deleterious variants that were there were more deleterious in people who had SUDEP than in people who hadn’t.
So, we found two things – that there were more of these variants and the ones that were there seemed to be, on average, more deleterious. It’s not one particular gene or another particular gene; it seems to be the sum across the genome. That was our key finding.
We did go on to look at whether there might be some genes that were overrepresented and when we did this, one gene called SCN1A did appear to be overrepresented. This is a gene that we know can cause a variety of epilepsies.
That leads us to question whether it’s the gene causing the severe epilepsy, which then increases the risk of SUDEP or whether the gene mutation has a separate effect, independent of the severity of the epilepsy. We can’t answer that question from our results and it’s possible the mutation might do both. It might increase your risk of SUDEP in the context of also giving you epilepsy, we don’t know.
Although it is speculation, we don’t think that it is just one gene or a set of genes, but that it is likely to be a combination of environmental factors and a polygenic risk.
What additional testing is needed to further understand the role genetics plays in SUDEP?
I think there’s a number of things needed. Firstly, it’s always important that our findings are tested by other groups, who might look at this model and other models. We’re not proposing that our model is the only model. It may well be that in sufficiently large cohorts, you might find specific genes that when mutated increase the risk of SUDEP.
I think the first step is to do more replication and more testing of all these different models and once that’s done and we have some robust results, then there’s likely to be a number of possibilities.
One is that there are specific genes which when mutated increase the risk of SUDEP. They may also contribute to the risk of epilepsy itself.
Another possibility (and they’re not exclusive), is that it does turn out that our findings hold up when tested further, and that there is a polygenic risk, so if you’ve got a series of deleterious variants, then the effects add up together and sometimes, when provoked by a seizure, that all leads to a fatal pre-terminal brain stem event which eventually leads to death. If these things turn out to be validated or replicated in further studies, then there are ways in which these findings might affect clinical care .
One is that one can start testing for either the genome-wide risk burden or for mutations in specific genes, so that it is possible to give people an idea of the risk that they may carry of SUDEP. Obviously, there’s going to be all sorts of ethical issues around that, but at the end of the day, SUDEP is an important area for many people with epilepsy and for carers. I think anything that can be done to identify whether people are at greater risk and then put measures in place that might reduce that risk is a good thing. We already try to do that where we can.
The second thing I suppose that you might do practically and from a research perspective, is to understand why these factors increase the risk of SUDEP. That will be more of a challenge because trying to make polygenic models is not going to be an easy thing to do, because we know the periphery probably needs to be involved in the causation of SUDEP, and in vitro models may not be sufficient.
How much research has been carried out on epigenetics and SUDEP?
Well that’s a very good question actually because obviously there are important modifiers of what genetic variation does. To my knowledge, it’s not been looked at. I agree that it is something that needs to be looked at because, of course, it could be one of the ways that environmental factors mediate risk with a specific seizure. However, I’m not aware that anyone’s actually done any work in that area.
Epigenetics is an area that people are just beginning to look at. I think a key problem there is of course that the epigenetics profile is often organ-specific and that’s obviously not something that’s easy to look at in epilepsy.
What do you think the future holds for patients with epilepsy? Are more personalised treatments on the horizon?
I think there are two broad approaches at the moment in terms of treatment and how we can improve that for epilepsy. One is a drug-centred approach, which is where we look at different drugs and how a large number of people respond to them and whether we can pick out people who don’t respond, people who do respond and people who get adverse effects.
Coupled with that, there is increasing interest and some early success being seen with what people call ‘precision medicine’ for individuals with epilepsy. What we’re seeing, especially with genetics, is that the epilepsies are super-segmenting. Clinicians have long recognised that there are lots of different types of epilepsy, but they haven’t really been able to quantify that. Genetics is underpinning these clinical observations and showing that there are a large number of rare conditions sharing the common feature of seizures, and these rare conditions all together contribute importantly to the common disease, or condition, of epilepsy. That is to say, epilepsy is not one disease, but many.
If we can begin to understand those super segmented phenotypes, then we will probably find that these will have their own best treatments, which may be existing drugs, new drugs or repurposed drugs. I think that is the way that it’s going to go for many more people than it does currently. Better understanding will lead to more personalized treatment, even if it is not ‘precision’ treatment.
Do you think it will be possible to prevent SUDEP altogether one day or is that unrealistic?
I don’t think we know enough about it to say! We should certainly be aiming to find out.
It may be that as our understanding increases, we will find that there are random or unpredictable factors that mean it’s impossible to prevent it completely, but I think there are things that are moving ahead that will hopefully reduce risk for more people.