Dr. Ching Soong Khoo: . [00:00:00] Hi everyone. Welcome to the ILAE Sharp Waves podcast. My name is Ching Soong Khoo. I'm a neurologist based at the National University of Malaysia. Today we are going to discuss an important topic, which is acute symptomatic seizures. I'm delighted to be joined by Marian Galovic. Hi, Marian. Could you please briefly introduce yourself? 

Dr. Marian Galovic: Hi. Thank you for the very kind introduction. I am an adult neurologist and epileptologist based in Zurich in Switzerland. And I'm leading the epilepsy unit at University Hospital in Zurich, and I have a strong interest in acquired epilepsy, epilepsy in older adults and also acute symptomatic seizures.

Dr. Ching Soong Khoo: Perfect. Let me start. What is the size of the problem of acute symptomatic seizures and how common are they? 

Dr. Marian Galovic: You know, acute symptomatic seizures are actually very common. They're probably more common [00:01:00] than most people think. If we look at all first seizures, about a third will be acute symptomatic, and if we also count febrile seizures, we're talking about roughly half.

So this is not a rare problem at all. To put that into perspective, about 1 in 20 people will have a seizure at some point in their life. That means at least 1 in 60 or 70 people will have an acute symptomatic seizure, and that's quite a lot. Yet, despite that, these seizures often just get a few lines in textbooks.

Most of us are trained to manage epilepsy but not really symptomatic seizures, which is a pity, because every neurologist working in emergency room on a stroke unit or in an intensive care unit will see them regularly. And they can happen in so many different situations – after stroke, head injury, during severe hyponatremia, inflammation like encephalitis [00:02:00] or even after alcohol or benzodiazepine withdrawal.

So they really are part of everyday neurology. 

Dr. Ching Soong Khoo: I think the next important question would be, what is the difference between an acute symptomatic seizure and an unprovoked seizure? 

Dr. Marian Galovic: That is a great question because the two can look exactly the same, but they are caused by very different mechanisms.

So in epilepsy, there is an existing epileptic network in the brain and this network generates seizures spontaneously, so without any external trigger. And this is a tricky bit about epilepsy, because seizures can happen spontaneously and without really us knowing when they are going to happen and when they're going to occur.

So in acute symptomatic seizures, the situation's quite different. There's always a trigger, which is an acute brain insult. And this could be a structural brain insult like an acute stroke or a traumatic brain injury, or it could be due to [00:03:00] inflammation of the brain, or it could be due to a metabolic or toxic insult of the CNS.

So even a person with a previously healthy brain can have a seizure if the trigger is strong enough. Now, let me give you an example. So, if somebody has very low sodium, called hyponatremia, the brain will swell and this may release excitatory neurotransmitters like glutamate. This can then provoke a seizure. But once the sodium normalizes, this risk disappears.

Or take a stroke. So in the acute phase when there was an acute stroke, there will be some inflammation, there will be cell death, release of excitatory neurotransmitters, disruption of the blood brain barrier, and all of these factors can trigger seizures.

But over the next few weeks, things calm down and, if then a seizure starts months later, months after a stroke, that's when we are usually dealing with an unprovoked seizure, then that is called [00:04:00] post-stroke epilepsy, and that is a completely different process. But in the acute phase, these seizures are triggered. So even though they look similar, the underlying mechanisms are very, very different. 

Dr. Ching Soong Khoo: Thank you so much for addressing this very complex topic, per se. So with the current definition, do you think there are any critical issues and potential pitfalls? 

Dr. Marian Galovic: So overall, the current definition works really well. It's very clear and practical, but yes, there are a couple of tricky points. So first, the current definition is intentionally very specific. So for example, for metabolic disturbances, the cutoffs are set at the extreme end of abnormal. So that means we might miss some people with milder abnormalities who still have a seizure because of them. And clinical judgment is really key here, and especially also how fast the disturbance develops can matter more than how severe [00:05:00] it is. So if hyponatremia develops very quickly, this is more likely to trigger a seizure. 

And secondly, there's the issue of timing. So for structural causes like stroke or trauma, we use a seven-day cutoff to define acute symptomatic seizures. But does a seizure on day seven really differ from one on day eight? Probably not. So the evidence for this cutoff is so far quite limited, and this temporal cutoff is also quite arbitrary. So that is why in the SeLECT Consortium, that is a large consortium researching post-stroke epilepsy, we currently looking at thousands of cases, and what we can see is that seizures during the first few weeks, not only the first week, but during the first few weeks, maybe up to one month, seem to behave like acute symptomatic seizures. So in other words, they have low risks of later unprovoked seizures. And only after that, the risk of unprovoked seizures rises. So that means that [00:06:00] probably only seizures occurring more than one month after stroke can be reliably classified as unprovoked. 

But the situation is even more complicated if the underlying condition persists. So a typical example is encephalitis, either infectious encephalitis or autoimmune encephalitis. So in these cases, there will be a risk of acute symptomatic seizures as long as a sufficiently active inflammation in the brain is present. So thus it may be really difficult to determine when seizures should be considered as acute symptomatic, and when there may be the transition to unprovoked seizures and actually epilepsy.

And finally, not all acute symptomatic seizures are really equal. Seizures due to structural causes, such as a stroke or trauma often predict later epilepsy. So they are associate with a high risk of later epilepsy. In contrast, metabolic or toxic causes, such as low sodium or drug effects almost never cause, or are almost never associated with later [00:07:00] epilepsy.

And if the acute symptomatic seizure presents as a status epilepticus the risks for developing epilepsy are much higher and also the risk of mortality is much higher.

Dr. Ching Soong Khoo: I see. So do acute symptomatic seizures have an impact on outcomes such as mortality, and is it etiology specific? 

Dr. Marian Galovic: Yes, very clearly they have an impact. So people who have acute symptomatic seizures tend to have worse outcomes, and part of that is because they usually have more severe brain injuries, so that seizures are marker of a bad injury. And poor outcomes are also in these cases mainly driven by the underlying etiology, such as a stroke, infection, or trauma. 

But there’s also growing evidence that the seizures themselves may be making things worse. They are increasing metabolic demand and by that they can worsen secondary brain injuries or they can cause, actually, secondary brain injuries in people who already have an [00:08:00] acute brain injury. And particularly bad in this regard are prolonged seizures and status epilepticus. So these can cause excitotoxicity and in this manner, they can directly damage neurons.

In one of our studies that we did after stroke, acute symptomatic status epilepticus was by far the strongest predictor of mortality. It in increased the risk of mortality by 13-fold, even after adjusting for many, many other covariates and confounders. So yes, these seizures seem really to matter.

Dr. Ching Soong Khoo: So in your opinion, when do acute symptomatic seizures become epileptic seizures? And are there any tools that can help us to prognosticate. 

Dr. Marian Galovic: So I would, I wouldn't actually say that they become epilepsy because the mechanisms are fundamentally different. But having an acute symptomatic seizure, especially after structural injury such as a stroke or a trauma, is a red flag. So it means that the injury was severe, [00:09:00] it most likely affected the cortex, and probably tapped into a brain network that is likely to generate seizures. 

I also believe that acute symptomatic seizures are in a way a marker of an individual's intrinsic seizure threshold. So this is something that may be influenced by many factors, also including genetics. In other words, a brain with a naturally lower seizure threshold is more likely to generate an acute symptomatic seizure in the context of an acute brain injury. But this brain is probably also more likely to generate seizures later on during epileptogenesis, and also once epilepsy has developed. So acute symptomatic seizures may actually serve as an important marker of the brain's seizure threshold or so to speak, of an ictogenic brain injury, and are therefore associated with a heightened risk of developing epilepsy.

So I see them as a mark of a brain that's prone to develop epilepsy. 

Coming back to your question, here's, here's how I think about it. [00:10:00] Seizures in the first week after an underlying brain injury almost always can be considered as, as acute symptomatic. Seizures about one month after brain injury are almost always unprovoked. And the tricky period is in between weeks two to four where we really have to be cautious before labeling it clearly as epilepsy and starting long-term treatment, and we need to be more cautious in this regard. 

And so the question was also whether there are any tools that can help prognostication? Yes, there are a couple of tools regarding this. We have been developing within the SeLECT Consortium the so-called SeLECT model, which is a validated and very easy-to-use model to predict the risk of seizures and developing epilepsy after ischemic stroke.

There's another model that's very commonly used for hemorrhagic stroke called CAVE. And then there are a couple of other models for other conditions like subarachnoid hemorrhage, cerebral venous thrombosis, and it may get quite overwhelming to have all these models in all different [00:11:00] places and to use them.

So what we also created quite recently is an online calculator or website that manages all these calculators. And the website is called predictepilepsy.com. Definitely do try it out. It's very simple to use and it gives you all the background information on all these models and tools. And I think it's easy to use and maybe helpful for counseling patients with a brain injury and acute symptomatic seizure about risk of developing epilepsy. 

Dr. Ching Soong Khoo: How are acute symptomatic disease managed? Are there any guidelines for management and what is done in real world situations?

Dr. Marian Galovic: Yes, there are some guidelines, but the evidence space is really thin. So surveys show that 90% of neurologists really feel uncertain about how to best treat these acute symptomatic seizures and that's completely understandable. For me, the main thing to remember is that acute symptomatic seizures should not be managed like epilepsy. [00:12:00] Epilepsy means long-term risk, so long-term treatment makes sense. And treatment in epilepsy is usually started slowly and gradually until a sufficient effect is reached. And then treatment is maintained typically at least for a couple of years. 

The management of acute symptomatic seizures is essentially actually the opposite. So first the clinician actually needs to decide whether the seizure actually requires treatment at all. Many guidelines recommend not treating acute symptomatic seizures because they generally have a relatively low risk of recurrence and a relatively low risk of late epilepsy. 

There is a clear disconnect between guidelines and reality. So in practice, 80% to 90% of acute symptomatic seizures are actually treated with antiseizure medications. Now, why is that? So in conditions like acute stroke, seizures can cause falls and injuries and they can also increase the brain's metabolic demand. There were good studies [00:13:00] showing that during a seizure in neurocritical care, there's increased oxygen consumption of the brain, and there's increased metabolic demand of the brain. And if you already have an injured brain, this can lead to secondary neurological injury, secondary neurological worsening and this is what many physicians are rightly concerned about. This is especially relevant when perfusion is reduced. For example, if you have a patient with large artery stenosis or vasospasms. So on the other hand even acute symptomatic seizures due to non-structural metabolic or toxic disorder may need short-term antiseizure medication treatment.

And that's quite logical. If you have a patient with severe hyponatremia or a patient who has alcohol withdrawal, there remains a high risk of recurrent acute symptomatic seizures, or even acute symptomatic status epilepticus as long as the metabolic derangement or the withdrawal persists. So in these cases, many clinicians will consider brief course of antiseizure medication [00:14:00] while the provoking factor remains.

So the next step, if a physician decides to treat actual acute symptomatic seizures with antiseizure medications, how does the management then differ from epilepsy? In acute symptomatic seizure, you want to hit early, fast, and hard, and that is because you don't want to wait until you sort of slowly increase your medication to an efficient dose because your provoking factor is there.

So you want to act quite quickly and you want to use fast-acting medications at an effective dose. And you want to do this during the acute phase. And then after the acute phase, you want to taper off once the underlying issue has stabilized. So in most cases I'd recommend treating from somewhere between one week to around three months and rarely longer unless you have very, very good reasons to continue treatment long term. For example, a very high risk of recurrence predicted by these models that we've discussed before. [00:15:00] 

In some places such as the United States, I know that physicians may prefer to keep treatment for around 6 or 12 months, and that's all right. But ultimately, all experts agree that long-term therapy is usually unnecessary for acute symptom seizures unless you can clearly demonstrate a very, very high risk of further seizures or the development of epilepsy.

So, in other words, with acute symptomatic seizures, you either don't treat at all or you treat hard and fast and then withdraw treatment quickly once the acute seizure risk has passed. And so this treatment approach is very, very different to what we would do in epilepsy with unprovoked seizures. 

Dr. Ching Soong Khoo: Following up on the question—are there any preferred anti-seizure medications? 

Dr. Marian Galovic: Yes, and, and there is very good survey data showing that around 90% of acute symptomatic seizures are treated with levetiracetam, and the rest usually get lacosamide or valproate. And the reason is that these are all fast-acting medications. They can be given [00:16:00] intravenously and they are generally very safe.

Valproate can be slightly more tricky because it has interactions whereas levetiracetam and lacosamide are relatively safe regarding interactions, but it's still used in some cases where, for example, levetiracetam and lacosamide are not great options. So for short-term treatment where we need to act quite quickly, levetiracetam and lacosamide usually are great first-line choices. 

Dr. Ching Soong Khoo: Does treating acute symptomatic seizures with ASMs versus not treating affect outcomes, and are there any data on, in-hospital only treatment versus longer-term treatment? 

Dr. Marian Galovic: Yeah, well that's the million-dollar question, and so far the answer is we don't have strong evidence that treatment really improves outcomes.

So while acute symptomatic seizures clearly are linked to worse outcomes, studies so far haven't reliably shown that treating them actually [00:17:00] improves recovery of these brain injuries like stroke or trauma. 

When it comes to treatment duration, data suggest that prolonged treatment offers no extra benefit. So again, one week to three months is usually enough and there’s no evidence that treating long actually provides any benefit. But having said that, most of this data is retrospective and may really be confounded, because more severely affected individuals may be more likely to receive treatment. So really making a final call is really difficult at the moment, and we really need prospective trials to look into this.

And in some very well selected cases, longer treatment with antiseizure medications may indeed be helpful. Specifically, in patients with a very high risk of developing epilepsy after brain injury. So as I mentioned, there are several excellent validated and easy-to-use models to predict this risk, such as the SeLECT score for ischemic stroke or the CAVE model for intracerebral hemorrhage. [00:18:00] And if the predicted risk exceeds about 60% within the next 10 years, some clinicians will consider offering longer-term antiseizure medication treatment. 

But of course in this regard and in general, this is shared decision making. So you always need to involve the patient, discuss the potential benefits and risks with them, and really make the decision together. Particularly when there was no unprovoked seizure, you don't really have the diagnosis of epilepsy, but you're really concerned that the risk is very high, that there may be further seizures and that epilepsy may develop and you would want to start prophylactic treatment in this regard. 

Dr. Ching Soong Khoo: Great. So what about anti-epileptogenesis? Could treatment with ASMs reduce the risk of developing seizures or epilepsy in the future? 

Dr. Marian Galovic: This is a really exciting field right now because traditionally we believe that antiseizure medications suppress seizures, but they wouldn't change the [00:19:00] cause of the disease. But there are now some emerging hints that certain drugs might actually have an anti-epileptogenic potential. So that's a completed study using eslicarbazepine for one month after stroke, and there's currently another ongoing study using perampanel for 12 weeks after stroke. 

So the eslicarbazepine study was stopped early because of recruitment issues during the COVID 19 pandemic. So it was slightly underpowered. So the trial, ultimately it failed to meet its primary endpoint. When you look more closely at the study data, there appears to be a large and clinically meaningful effect of one-month eslicarbazepine treatment in reducing the risk of post-stroke epilepsy. And this is really amazing.

So in the eslicarbazepine group, the risk of post-stroke epilepsy was reduced by about 50%. And although this was non-significant, it seems to be a clearly clinically relevant effect. [00:20:00] And probably we need just a larger sample size to reach statistical power here. And also the number needed to treat was really encouraging, so the number needed to treat to prevent one case of post-stroke epilepsy with this one-month eslicarbazepine treatment was between 12 to 16, which I found is quite remarkable. So clearly further studies need to explore this in more detail and in a fully powered cohort. But these results are really exciting and encouraging. 

And we are also watching this ongoing perampanel study because I heard that interim analysis may be showing similar effects, which would be quite amazing.

So I think the coming years could bring quite a wealth of new evidence. Maybe a paradigm shift. So if some medications show antiepileptogenic effects, we may be considering using them not only to treat seizures or to suppress seizures, but also to prevent a development of epilepsy, which really would change how we think and treat [00:21:00] epilepsy.

Dr. Ching Soong Khoo: What is the current evidence on giving prophylactic ASMs for certain conditions like traumatic brain injury, hemorrhagic stroke, and so on? And if you decide to give, how long should we treat? 

Dr. Marian Galovic: So, I'm generally quite skeptical about giving antiseizure medications prophylactically, meaning before an acute symptomatic seizure even happens. So unless you have really clear evidence of a high risk, for example, if, if you see some good evidence on an EEG or something similar, unless you have that, it's usually not justified. 

I think in those who do have acute symptomatic seizures, short-term treatment is usually fine. Again, around one week to three months is probably useful, but long-term prophylaxis doesn't seem to help.

So it's really important to see patients again in clinic to discuss this with them. Maybe do another EEG, do these scores calculate the recurrence risk of unprovoked seizures and the risk of developing epilepsy using [00:22:00] these scores that I've mentioned. And then really most cases to withdraw treatment because this is not necessarily long term in most cases.

Dr. Ching Soong Khoo: Talking about EEG—should EEG be routinely performed in conditions highly associated with acute symptomatic seizures? And are electrographic-only acute symptomatic seizures treated the same as clinical acute symptomatic seizures?

Dr. Marian Galovic: So this is a great point because the EEG is incredibly useful here. And in acute brain injury, we know that many seizures are electrographic only. That means that you wouldn't see them clinically and you would otherwise miss them. And there was one very well-done study that showed that more than 40% of patients with intracerebral hemorrhage had purely electrographic seizures. So without EEG, we would miss these electrographic seizures in 40% of these patients.

So in the context of an acute brain injury or in neurocritical [00:23:00] care, EEGs are extremely helpful for really peeking below the surface and detecting seizures that you would otherwise not notice. 

Now, the other question is, what's the optimal duration of EEG monitoring? Is this really a critical factor? Because longer monitoring will increase the likelihood of detecting seizures. And in this regard, the local availability of continuous EEG monitoring will really determine how many patients you can actually offer this to. And there are tools like the 2HELPS2B score and other predictive models that can also help you identify high-risk patients who are most likely to benefit from prolonged EEG monitoring.

And the question was also how relevant are electrographic seizures, and this remains somewhat contentious. But if we ask about this in surveys, we see that most clinicians will treat them very similarly to clinical seizures. So most clinicians will decide to treat electrographic [00:24:00] seizures at least as aggressively as they would treat clinical seizures.

And that makes sense, because there is some consistent evidence that points to the prognostic importance of electrographic seizures and the electrographic seizure burden, as this seems to be associated with worse outcomes. So another important observation is that if you see one electrographic seizure on EEG, it's very likely that there will be more than this. So honestly, I can't recall a single case where I only saw a single electrographic seizure. So when we see one electrographic seizure on the EEG, that's almost always going to be more seizures happening. So in these cases, really prolonged EEG monitoring is quite helpful. 

And I do believe that EEG is not just diagnostics, not just telling you, well, are there seizures or not, but it's also prognostic. And in our recent work we found that detecting epileptic abnormalities, such as ventricular discharges or these rhythmic periodic patterns or electrographic [00:25:00] seizures, on an early EEG which is performed early, so in this first seven days after stroke, that this is actually predictive of developing late epilepsy.

So our results indicate that the prognostic value of these EEG patterns is similar to that of an acute symptomatic seizure. And that makes sense. So these epileptiform EEG findings, they probably reflect a more ictogenic or epileptogenic type of brain injury, helping you to predict who is at high risk of developing epilepsy.

And this is particularly useful in patients who did not have an acute symptomatic seizure, where the EEG can instead reveal this highly ictogenic effect of the stroke. And we've also built this into the newest SeLECT model. And again, you can find it on predictepilepsy.com. It's a great tool and I really encourage you to check it out, both for research and also when counseling patients about these long-term risks.

Dr. Ching Soong Khoo: Are there any points or thoughts that you'd like to share before we end this [00:26:00] session? 

Dr. Marian Galovic: So I'm really happy that we're talking about acute symptomatic seizures because I do think that that frequently they are underrecognized, overtreated, and really they are not mentioned enough in textbooks and in lectures we give to medical students.

And so overall I think acute symptomatic seizures are common. I think they're really important, and quite underappreciated. And I think they do tell us a lot about the injury, about the brain's excitability, and also about who's likely to develop epilepsy later on. So understanding and managing them well can really make a difference in patient care, and that's why I do believe it's important to talk about them.