ISC 2026 | Risk factors and prediction tools for post-stroke epilepsy

If you look across studies, what you notice is that most of the risk factors for post-stroke epilepsy are factors that are directly related to the stroke. Some of the most important ones, I would say, the location of the stroke. Any stroke that is cortical versus subcortical carries an increase of post-stroke epilepsy. Similarly, a severe stroke also linked to an increase of epilepsy, which can be assessed using a scale such as the NIHSS scale. The other important risk factor is any, it’s not directly, I would say, if you have an early seizure, that actually multiplies by seven, your risk of having post-epilepsy. So having an early seizure, it’s a red flag that you may go on and develop late-onset epilepsy. Other factors include things like hemorrhagic transformation. When you have an ischemic stroke that turns into a bleed, that will increase your risk of developing post-stroke epilepsy. And not to forget to mention the fact that in some studies, it’s been shown that age plays a role, a large risk of disease plays a role as well. And more recently, having microbleeds, that is preexisting microbleeds, which are tiny bleeds, chronic bleeds in the brain, increase the risk of having late-onset epilepsy by two to three fold. And there have been models developed to predict post-stroke epilepsy. One of the most popular models is what we call the select model that was developed in Switzerland and published in 2018, which uses five parameters to predict post-stroke epilepsy. The S stands for, I think S is the severity of a stroke and you have early seizures as well as one of the variables in the model. Large actual disease is one of the variables that account in the model. Cortical location is one of the variables. And the last one is the MC33. So all those factors combined will help providers or physicians predict the risk of late-onset seizures or post-stroke epilepsy. Now, having said that, the model has been developed in several cohorts and has been updated to include variables such as early status epilepticus. There’s one other update that was made that includes electroencephalographic variables as well. This is a select model. Beside the select model, and more recently, there’s one model called the ischemia model that was developed by a team at Harvard. And what they did in that model that was different from what the select model did was to include a variable that will account for the hemorrhagic transformation. Remember, we said earlier that having hemorrhagic transformation has an increased risk of developing late seizure, that’s significantly. The next step would be to integrate variables such as the patient’s genetic predisposition or include variables such as cerebral microbleeds. But these are two models that we currently use to predict late-onset seizures after ischemic stroke. After hemorrhagic stroke, there have been models that were developed and the one that is probably the most popular is called the CAVE model. The CAVE model needs to be improved and we really need to develop new models for post-hemorrhagic stroke epilepsy.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.