Sharp Waves: ILAE's epilepsy podcast
Sharp Waves: ILAE's epilepsy podcast
Drug-resistant epilepsy: Dr. Patrick Kwan
About one-third of people with epilepsy will continue having seizures despite medications. What is the clinical definition of drug-resistant epilepsy? Why does it occur, and is it possible to predict? Dr. Cecilie Nome interviews Dr. Patrick Kwan.
Relevant papers:
The costs of epilepsy in Australia | Neurology
Sharp Waves episodes are meant for informational purposes only, and not as clinical or medical advice.
Let us know how we're doing: podcast@ilae.org.
The International League Against Epilepsy is the world's preeminent association of health professionals and scientists, working toward a world where no person's life is limited by epilepsy. Visit us on Facebook, Instagram, and LinkedIn.
Dr. Cecilie Nome: Okay. Everyone, welcome to this podcast episode. My name is Cecilia Nome and I will be your host for this episode. Today, we will dive into a very important and challenging topic, and the topic in question is drug-resistant epilepsy. It affects quite a lot of people with epilepsy and it can have some profound impacts on their family life, independence, and quality of life. To tackle this important topic, we are honored to welcome professor Patrick Kwan. Would you maybe like to introduce yourself?
Dr. Patrick Kwan: Sure, of course. So my name is Patrick Kwan. I'm a professor of neurology at Monash University in Melbourne, Australia. I'm also a physician. I'm a neurologist working mainly epilepsy at the Alfred Hospital, also in Melbourne, Australia.
Dr. Cecilie Nome: Thank you so much. I think we should jump right into the definition of [00:01:00] drug-resistant epilepsy. Could you define this for us?
Dr. Patrick Kwan: Well, it depends on who you ask this question and what version of the definition, I guess. So currently drug-resistant epilepsy is defined by an ILAE task force; that was published back in 2010. In that definition, drug-resistant epilepsy is defined as an epilepsy condition that has failed to respond to two anti-seizure medications that have been taken at adequate doses and durations based on the clinician's assessment and that these medications are appropriate for the patient’s particular epilepsy syndrome or the seizure type.
Dr. Cecilie Nome: So does it matter which types of anti-seizure medications you try? Or does is it like as long as it works for the type of [00:02:00] epilepsy the patient has then it's fine.
Dr. Patrick Kwan: What constitutes appropriateness? That's a very good question. So the concept and the definition is that it should be appropriate for that particular type or put it the other way around, there should have been evidence that that particular medication should be effective for that particular seizure type of syndrome. So that could be coming from clinical trials. But equally then, the drug should not be inappropriate for the seizure type because we know some medications can exacerbate certain types of seizures.
The community is conscious that the current definition may not encompass all scenarios, because as you can appreciate it is a fairly simple concept. You know, two drugs, you know, two strikes you’re out, kind of. And when it was proposed at the time, the group was very conscious to have something that is easily applicable.
We know that [00:03:00] 60% to 80% of people with epilepsy live in resource-poor settings or countries who may not have all the sophisticated technologies to do whatever investigations that may come up and in the future. So we want something that is actionable, and that could also act as a pointer or a trigger for specialist evaluation. And with that in mind, it is a very clinically oriented definition. It wasn't intended to be applied outside the context. It wasn't a definition for, for example, for someone developing animal models for drugs in epilepsy, or a cellular model or preclinical model.
The primary concern was a clinical definition that applies in both children and adults, but the intention is that it can be modified and built upon depending on the actual scenario. The International League Against Epilepsy has formed a new task force to look into it and to [00:04:00] update or revise or supplement the current definition.
I'll make one more point if I may. The reason for choosing the two medications is we know from epidemiology studies that when patients are in that situation, when two medications fail, their chance of becomes seizure free is greatly diminished compared to someone who's drug naive. And again, that is a relative concept, and reduced doesn't mean zero. So again, the patient may still become seizure free, but also, we need to be conscious that epidemiology evidence comes from a body of evidence. And most of the studies done at the time was using fairly old methodology.
Obviously, science has improved and has moved along, and we now have tools that may allow us to do individualized prediction. And talk about artificial intelligence here, of course, where it can come up with individualized prediction and allow more personalized and more nuanced assessment. [00:05:00]
So I think there is an argument that when we update the definition, revise it, we take into account the predicted probability of seizure freedom, not just observed. And maybe they can supplement each other. The reason being that at least in a hypothetical situation, if an individual can be predicted to have a very low chance of becoming seizure free, then should the non-drug treatment be considered much, much earlier at the outset, even before the patient has been observed to fail the two medications.
Because for two medications to fail to work, that means the patient will have to endure seizures. They may have to endure the side effects on the medications. And when you're faced with a non-drug intervention that may have a higher chance to be effective, then I think it raises the question, should the non-drug therapy be used [00:06:00] earlier, maybe even together with the medications, rather than waiting until two medications have failed.
Dr. Cecilie Nome: That makes a lot of sense. And I'm, I'm really looking forward to this new updated definition because it'll be, I mean, things have definitely changed since the first one was made, so that will be very interesting.
Often in studies and, and books about epilepsy, we do read that about one-third of patients don't respond to medication. Would that still be true, you think, or?
Dr. Patrick Kwan: Hmm. Yeah, that's a good question. So if you look at studies that really see you know, what is the proportion of patients fulfilling the actual definition, I think you'll find a range. I think most of you are shown between maybe 25% to one-third, 20% to a third, that sort of range. So I mean, I think it's not 90%, but it's not 10%. So it is a substantial proportion of patients.
Dr. Cecilie Nome: And just to follow up on that, are there any [00:07:00] differences between different age groups or different countries regarding how many patients with epilepsy that develop drug resistant epilepsy?
Dr. Patrick Kwan: Hmm. Yeah, so there's been quite a bit of work looking at some of the other facts and the factors as you mentioned.
So, in terms of age group, generally speaking, you know, pediatric onset epilepsy seems to have a better outcome than the other onset ones. However, there are very severe childhood epilepsy syndromes, the developmental and epileptic encephalopathies are very drug resistant.
But if you look at age group as a whole, then it seems to be the young age group have better outcomes in general. And I'm not aware of any study directly comparing different countries, although there is ongoing research in this area, including some that we are doing ourselves with collaborators around the world, looking at differences in treatment response [00:08:00] across different countries and continents.
The type of epilepsy is very important, of course. So focal epilepsy tends to have worse outcomes than generalized epilepsy. Again, this is on a very general level. You know, idiopathic generalized epilepsy tends to be a bit more drug responsive, but then there are also patients with IG or when the IG is very drug resistant as well. We see those cases too.
I think partly because a lot of the syndromes are defined phenomenologically and supported by neurophysiology of course, but not by the underlying etiology. Some we know the genetics, many we don't. So these are still heterogeneous conditions.
Dr. Cecilie Nome: So when I was preparing for this podcast, I also came across the term “pseudo drug resistant epilepsy.” And if I understand it correctly, this is like a misdiagnosis or that the patient has a different condition which is not epilepsy, but it's believed to [00:09:00] be epilepsy or misdiagnosed as epilepsy. Is this like a problem that you would face quite often or do you know?
Dr. Patrick Kwan: Yeah, no, that's a very good question. The term so-called pseudo drug-resistant epilepsy is meant to imply that not necessarily the patient doesn't have epilepsy, but that the epilepsy is perhaps not drug resistant, actually resistant to medication. There are other reasons the patient has not been responding to the medication. And one of the reasons, of course, is misdiagnosis. If the patient does not have epilepsy, then the antiseizure medications are not going to help them. But there could be other reasons: The patient was given, quote unquote, the wrong medication for the particular seizure type of syndromes, so they haven't been treated properly. Another reason could be the medication not being used at the right dose. So maybe the right type of medication [00:10:00], but not a sufficient dosage.
And then finally, there are reasons that are related to more lifestyle. We know that there are some lifestyle triggers for seizures that can also interfere or influence or impact the response to treatment or medication or adherence for whatever reasons. That can also lead to the patient appearing to be drug resistant when in fact it’s not because of the fundamental effectiveness of the medication, per se.
Dr. Cecilie Nome: Thank you for clarifying this. Are there any specific risk factors for drug-resistant epilepsy?
Dr. Patrick Kwan: There are certain syndromes or related to the underlying pathology, etiologies that are known to be more likely to be drug resistant. Best known examples could include hippocampal sclerosis or certain malformations of cortical development such as focal cortical dysplasia. And that's why in the literature you often [00:11:00] find this, well in practice as well, this being often patients requiring surgical treatment.
It's been reported that people with other mental health comorbidities such as depression, anxiety are also more likely to be drug resistant. And again, that could be multifactorial and it's a bit of chicken and egg situation, but also likely to be bidirectional. After all, we're talking about the same brain, and the neurotransmitters could influence both the excitability as well as transmitters related to the mood and so on.
So it's not difficult to understand why epilepsy can be influenced by comorbidities and how they may be associated with poor treatment response.
Dr. Cecilie Nome: It's also possible that some types of epilepsy, they have a tendency to become drug resistant [00:12:00] over time. For example, like hippocampal sclerosis, if you have temporal lobe epilepsy, that is, could develop. Is that correct?
Dr. Patrick Kwan: Mm. That is a debated concept. You know, does drug resistance develop over time, or is it de novo? Meaning that at the beginning, the onset, the epilepsy was already drug resistant and we are just finding out by trialing some medications and then find out they failed, or does it develop over time?
And I think the problem is we are sometimes using epilepsy as a starting point, whereas the epileptogenesis actually starts a lot earlier. Some say from the first seizure, but probably even from the first insult that led to the cascade of biological and genetic and functional structural changes in the brain as part of the epileptogenesis.
So. If you understand it this way, then drug [00:13:00] resistance is a continuum of the epileptogenesis and at what inflection point we define this patient now as drug resistant in some way is arbitrary, but it's operational. Because we need this operation definition to allow us to take certain actions.
But if you look at it, think of it biologically, this is a continuing process. And I would argue that the process starts even before the first brain insult because we sometimes, you know, as epileptologists or epilepsy researchers we sometimes forget that the patient actually, they have a life before they have epilepsy, before the first seizure.
something that we're very interested and we're looking at the pre-epilepsy or pre-seizure analysis, you know, are there predispositions of someone having even their first seizure? And so as to understand, shed new light on [00:14:00] understanding the whole mechanism of epilepsy.
Dr. Cecilie Nome: Do we know what type of pathological processes that is known or believed to be underlying drug-resistant epilepsy?
Dr. Patrick Kwan: There's been a few hypotheses. Again, this is on a high level, suffice to say we don't have a good understanding of drug-resistant epilepsy. And it's likely that there are multiple factors and they're different between different individuals and that more than one factor exists for a particular individual, just like epilepsy itself.
One is around the concept of efflux cell transporters being over-expressed at the blood-brain barrier. And these are well studied in cancer, oncology, and they became prominent in the epilepsy field more than 20 years ago, around that time.
There was a lot of attention because these are transporters at the [00:15:00] epithelial cells in the capillaries in the brain. And they act as exporters. So they export the drugs from the brain parenchyma back into the lumen, the capillary lumen. And so they limit the accumulation of the anti-seizure medication at the site of action, at the focus, if you like.
So that's the concept and that is supported by experimental evidence to show that these transporters are upregulated at the epileptic focus. And, and also there is a lot of in vitro and in vivo evidence that certain antiseizure medications, quite a number of them, are substrates for these transporters. There was even initially some evidence to suggest genetic variation that may affect the expression of these transporters may be associated with drug resistance in people with epilepsy. Although subsequent studies have been showing [00:16:00] very inconsistent findings for this particular theory, I think the criticism has been mainly that it means it's largely pharmacokinetic or local pharmacokinetic problem, and one can overcome it by just increasing the dose of medication.
The counterargument of that is that the medication may then affect other parts of the brain causing side effects or limiting the dose prescribed. And the inhibitors of these transporters have been tried as well, not very seriously, I would say, but without very dramatic improvement. And there are a number of these different transporters as well. So it's an area very hard to tackle.
Another main theory is around alteration in the target of the antiseizure medications. If you think of the main targets, these would include glutamate receptors, GABA receptors, or ion channels. And there is experimental [00:17:00] evidence and also evidence from human tissues that these targets are altered in people with drug-resistant epilepsy.
Another prominent theory has to do with the severity of epilepsy. This concept is the drug resistance is just a reflection of how severe the epilepsy is. And that goes back to what I said earlier, the drug resistance is a continuum of the epileptogenic process. I think the criticism for this theory is that how do you define severity and what makes something more severe, less severe? Severity is more a clinical concept in some sense.
I think ultimately, it's down to our poor understanding of ictogenesis and epileptogenesis. And we have drugs that, for example, act on sodium channels, and these are common mechanisms of drugs, but they're not the final common [00:18:00] pathway or the original starting event, so to speak.
So we interfere in the whole process of epileptogenesis and ictogenesis. The onset of a seizure is a cascade of events, so we're interfering with parts of it and maybe that's enough for some patients, whereas for other patients we need to intervene at multiple steps in this process. And we don't have the tools to do that and that's why we call it drug resistant. But that's just because our drugs are not good enough.
Dr. Cecilie Nome: Okay. So those were some very interesting theories and as you say there's also a lot of things we don't know, but it's this process and the drugs that we have available today, they may not target all the steps, but maybe some of the steps somewhere in the middle, in the cascade.
I also read a bit about inflammation and these theories about inflammation and how this could sort of worsen the process. What are your thoughts on this topic?
Dr. Patrick Kwan: I [00:19:00] think inflammation, there's no doubt it plays a role. I think the difficulty for the field has been to define exactly what we mean by inflammation. It’s a very all-encompassing sort of concept and how do you intervene? Something we discussed earlier about transporters and because some anti-inflammatory agents can also interfere with the drug transporters, and it's been at least experimentally to show that it can improve or reduce drug resistance in animals. Now is that through inflammation or is that through a knock-on effect on the drug transporters or other mechanisms we're not aware of?
So I think we have a challenge in in demonstrating interfering with inflammation can be directly targeting this particular pathway.
Clinically, as a clinician, I don't really care what mechanism is working. I do care, but you know, what I most care is that it [00:20:00] works. So I think there is a certain potential role to play in interfering with the inflammatory system. But I think we need to understand a bit more the nuances.
Dr. Cecilie Nome: Thank you. When I prepared for this episode, I was also bumping into blood-brain barrier dysfunction and also into a different topic, which was epigenetic factors. Do you have anything about these types of concepts?
Dr. Patrick Kwan: Mm-hmm. Mm. Yeah. I mean, blood-brain barrier. We touched on again briefly about the concept of transporters. So these transporters are, in fact, they contribute to the blood-brain barrier. But we also know that during seizures there is a breakdown in the blood-brain barrier. How and whether that can be harnessed, I think in treating epilepsy, is something that hasn't been fully explored.
Given that you mentioned blood-brain barrier, [00:21:00] another area becoming interesting is the perivascular space. Another name for it is the glymphatic system. And that is our brain’s natural, sewage, if you like, waste sewage system, when we get rid of the toxins and rubbish from the brain.
And we and other groups have done some work looking at comparing perivascular spaces between people with epilepsy or other etiologies that may lead to epilepsy. And we can find a difference in the amount and the size of these perivascular spaces between people with epilepsy and without epilepsy.
What role do they play? Would the abnormality lead to a buildup of these toxins which may become excitatory and contribute to epileptogenesis? I mean, these are the questions we are asking. There's a lot of literature around the association between perivascular space and neurodegenerative diseases, especially of Alzheimer's. [00:22:00] And we know there's a bidirectional relationship between Alzheimer's, the buildup of amyloid and, you know, tau not just in Alzheimer's but also in epilepsy. They also play a role as well. And most importantly I think for the listeners is to know that the perivascular spaces can be affected by poor sleep. So if you sleep well, then it helps the perivascular spaces to function and clear the toxins. So make sure you get a good night's sleep.
Epigenetics, this again, we're starting to look beyond just the coding genes in the genome. And, you know, epigenetics clearly, I think, plays a role because these changes from around environmental interventions and factors, looking at what else can affect gene expression, because it's not just about the sequence of the gene after it needs to be expressed. Apart from epigenetics, there are also other factors can affect gene expression such as transcription factors. There's also work being done looking at the role of transcription [00:23:00] factors in gene expression and how these can influence treatment response, both in a good way and also in a bad way, you know, side effects as well as efficacy.
Dr. Cecilie Nome: Thank you. That was a very good overview of different pathological mechanisms that can contribute to drug resistant epilepsy. Like you said in the beginning, it's most likely it could be more than one of these factors going on. Maybe everything is going on at the same time in the brain.
So are there any of these factors that is close to being developed into a new treatment for drug resistant epilepsy, or are there any ongoing studies?
Dr. Patrick Kwan: I'm not aware of any interventions, against the epigenetic profile or transcription factor profile directly at this stage. Although these are currently, I think, mostly being used as biomarkers, which [00:24:00] still would be clinically very useful if we can these identify more biomarkers for drug resistant epilepsy, then one may intervene earlier and consider non-drug treatments.
So the biomarkers, I think is a very hot topic. I come back to AI. Because we traditionally, we look at biomarkers one by one or in a panel, we’re very limited by sample size and the statistics. But with AI it gives us the power to look at it in a completely different way, very agnostic way. And you know, the listeners may have heard of supervised learning and unsupervised learning AI. And with unsupervised learning, the AI can find patterns that may not be obvious in a traditional statistical modeling. So I think AI is going to change how we approach biomarkers and how we apply them.
Dr. Cecilie Nome: So I would like to go back a bit and [00:25:00] if you now pretend that I'm a clinician, I am seeing a patient with epilepsy and I have tried two different anti-seizure medications, and the patient still has seizures.
What would be my next steps? How should I treat this patient?
Dr. Patrick Kwan: Yeah. So that's exactly the purpose of having the definition is to trigger those questions. And in that scenario, as you say, on at least on face value, then the patient would seem to tick the box.
We discussed earlier the first thing to make sure is the patient doesn't have pseudo drug-resistant epilepsy. And by going over, you know, considering all the different possibilities we discussed earlier about misdiagnosis or misclassification of the seizure type of syndrome, the right dose [00:26:00] being used and other lifestyle factors.
I think that is the number one. And that may sound trivial, but in fact as we all know misdiagnosed epilepsy is common. And if you look at video EEG monitoring units anywhere between 25% to 30% of patients when they leave the unit, they have a non-epilepsy diagnosis. And so they all went in assumed to have drug resistant epilepsy because they've not been responding to anti-seizure medications. But when they leave, a different diagnosis has been found. So it is actually very critical. So in that scenario that's why the emphasis is referring to a special center that has the facilities and expertise to undertake this kind of diagnostic assessment using what we still consider the gold standard, which is video EEG monitoring.
Once that's cleared, it's worth kind of starting to have that conversation with the patient. So I would usually [00:27:00] start introducing the concept of non-drug treatments. Because often patients, they haven't been introduced to the idea of surgery, for example, or other interventions and those are major undertakings for patients, you know, to have a surgery.
So it's good to introduce that concept early. In fact, I would usually introduce that concept at the diagnosis, and I would say that there are these options and give them a general idea. But the same time you start look also looking at are there other medications, for example. But you start a conversation around other strategies and not just medications.
Dr. Cecilie Nome: Thank you. Are there any specific considerations that clinicians should keep in mind when we treat this patient with drug resistance?
Dr. Patrick Kwan: I think understanding the patient's needs and expectations are very important. And keeping in mind we do not have a [00:28:00] magic bullet for every single patient. So every patient, their strategy should be individualized and is a partnership with the patient. Any intervention we give, medications, surgery, diet, neuromodulation, the patient may have different preferences and questions about each of these interventions because they are so different.
So, in a sense, in epilepsy we're blessed to have so many different choices of treatment. But on the other hand because they are so different in terms of their efficacy, harm, and risk and side effects, then it's a very nuanced discussion with the patient and you need to understand what their hopes and concerns are to ensure that you meet their needs, that they will come along on this journey together with you as the treating neurologist.
And I think also it's important to work in a team because [00:29:00] we know, as you know, we discussed that 70% of patients are drug responsive. The epilepsy is drug responsive, so there's something different about this 30%. And at the point of failing two medications, that is a critical juncture. And is important to be I think conscious of all our limitations and so working with your colleagues, with the teams, radiology, surgeons, your psychologist, psychiatrist, nursing, allied health – work as a team to make sure that we give the patient the best chance to response and to improve in their long-term outcome.
Dr. Cecilie Nome: I would like to take a step back and to look, I mean, when you look at drug-resistant epilepsy from a populational based view or from a country based view, you do realize that, there are quite a lot of costs actually. There was this study from Germany that estimated that one patient with drug resistant epilepsy would actually cost something between 12 and 14,000 euros per year.
Is there anything we can do to sort of help out with this?
Dr. Patrick Kwan: Hmm. Yeah. I mean, this is a very big impact on the society, on the societal level.
And I think sometimes we do forget that. And when we measure costs we also often just focus on the direct medical costs and the treatment costs, but there are other costs as well. And we've recently done some work looking at the cost in terms of productivity and how much, you know, GDP loss for the country when there's a sizable population that have [00:31:00] epilepsy that is drug resistant, and, and we've shown that there is huge productivity loss.
Because an epilepsy is condition that affects all ages and often in people in adulthood in their prime years of reproductive and working life. So it has a huge impact. But we show that with a simulation, if you could just improve the seizure freedom rate by just 5%, it could lead to billions of dollars of savings in terms of GDP for this patient population.
We hope that the message here is for any stakeholders investing in providing better treatment for epilepsy, and this includes healthcare providers, but also the industry or investors who are investing in developing new therapies. Please look at this because it makes a lot of economic sense. If you can just improve seizure freedom by 5% for such a small country like Australia [00:32:00] alone, it can save billions of dollars. So if you look around the world, the return of investment can be huge.
Dr. Cecilie Nome: Thank you so much. I think that was some very interesting perspectives. And I mean, 5% it, it doesn't sound that much and it's still so much money you could actually save from the economical point of view. But of course, also, these are people that have epilepsy have quite some impacts on their life. So for the patients too, this will be very important actually.
Dr. Patrick Kwan: And, and their family and their carers too.
Dr. Cecilie Nome: Definitely. Do you have any other comments, Patrick?
Dr. Patrick Kwan: No, I think we've covered a lot. Thank you, Cecilie.
Dr. Cecilie Nome: I think you have given us some very interesting and important information and thoughts about this. Thank you, Patrick, so much, for being here and to share all your knowledge.
Dr. Patrick Kwan: Thank you. It's been a pleasure.