Without a seizure there is no epilepsy. And conventionally epilepsy is defined as a predisposition to recurrent unprovoked seizures. So when is a single seizure enough to diagnose epilepsy? One such scenario is following stroke.

After a single late post-stroke seizure the recurrence risk is over 70%. Therefore anyone with a pre-existing stroke who suffers an unprovoked seizure automatically qualifies under ILAE criteria as epilepsy.

But how does stroke lead to epilepsy? The most salient aspect of a stroke is necrosis. Brain tissue dies during a stroke. And if the neurons are dead, how does a population of them become hypersynchronous?

One clue is to look at the actual seizure onset zones for folks with this type of epilepsy. The ictal onset and areas of high epileptogenicity were associated with remote as well as lesional and perilesional areas. Perilesional/extralesional implicates cortex not directly part of the stroke umbra but which somehow must have dynamics altered by the nearby infarct.

The pathophysiology of post-stroke epilepsy is also pretty interesting. In this review they implicate neurovascular unit dysfunction, neural network reorganization, genetics, and glial proliferation as mechanisms.

It’s important to note that all this is most pertinent to late post-stroke seizures—ie seizures later than 7 days post-stroke. Early post-stroke seizures are thought to result from different mechanisms and largely don’t predispose to further seizures although the ideal time cutoff is subject to revision.

(Also important to note: abbreviating acute symptomatic seizures as ASS is a GOATed move in a professional scientific paper.)

The SeLECT score is an interesting prognostic tool. Its utility appears to be early prediction of late post-stroke seizures based on the NIHSS, finding of large-artery atherosclerosis, presence of ASS, cortical involvement, and the territory of the stroke. The problematic factor is whether or not one wants to treat with antiseizure medications (ASMs) before any seizure has occurred, based on the SeLECT score.

On the one hand, prevention of seizures is an important good. Seizures are disruptive and sometimes fatal. But anti-epileptogenic treatments don’t currently exist (although studies are ongoing). And, as stated, there is no epilepsy without at least one seizure. So patients with high SeLECT scores who undergo treatment are choosing to undergo years of treatment, which may be associated with adverse effects, to prevent something which may have not happened. And then good luck prognosticating seizure freedom when they finally want to stop ASMs.