Strokes or Seizures? What's the Score?

[Box: see text].

Commentary

Stroke is a leading cause of death and disability in industrialized countries. Ischemic stroke has an overall prevalence of 2.6% in adults over 20, resulting in approximately 795,000 strokes per year. It is the fifth cause of death in the United States.[1] Stroke mortality has declined over time,[2] in part due to broader education targeting earlier symptom identification and greater awareness by the public and the health-care personnel regarding the possibility of time-sensitive treatment. In addition, improvements in interventions, allowing for longer windows of treatment, dedicated stroke units, and specialized hospital care have improved survival and outcomes. This results in a higher number of stroke survivors and with them the resulting late stroke complications, including physical and cognitive disabilities as well as poststroke epilepsy (PSE).

Seizures are a common occurrence at stroke presentation (1.53%-3%) and within the acute hospital admission (2.03%-14.6%). Seizures occurring at onset or within 7 days of stroke presentation are considered acute symptomatic seizures. Patients with cortical, extensive, or hemorrhagic strokes are thought to be at greater risk of developing acute seizures, but the risk attributable to each of these characteristics is not well established. The presence of acute poststroke seizures is associated with higher morbidity and mortality than patients with comparable strokes but no seizures.[3,4] The presence of prolonged acute seizures is also associated with the later development of PSE.[5,6]

Late seizures happen weeks to years following an acute stroke. Population cohorts with long-term follow-up described 6.4% to 11.5% of poststroke patients developed PSE with a cumulative risk over time, which continues to increase even as far as 10 years following the initial stroke.[6-8] Clinical markers of cortical involvement such as aphasia, visual neglect, or visual field cut along with stroke severity, particularly involving the MCA territory, are associated with a higher risk of PSE.

Merkler et al conducted an ambitious retrospective review which included 777,276 consecutive stroke survivor patients from 3 states over 8 years and 81,984 Medicare stroke survivors over 5 years. They found that 9.27% and 13.37%, respectively, of these patients developed PSE within 10 years. This represented a 7-fold higher risk of developing epilepsy than the general population. The authors found the highest predictors for developing seizures were: hemorrhagic stroke and having a stroke while young. In this large sample size, they were also able to describe an increased risk of PSE within the nonwhite stroke survivors, which was highest among the black population. Prior studies had identified younger age (<65) at stroke presentation as a predictor for the development of PSE.[7] These findings may be related to the presence of larger vessel strokes in the young compared to small vessel strokes seen in the older population.

Ethnic disparities in stroke risk follow a similar pattern with a higher stroke incidence among young and black population when compared to nonblacks. Death rates related to stroke are also higher in nonwhite populations.[9]

Galovic et al developed a prediction model to identify the risk of individual patients to develop PSE and validated it in 1169 patients form 3 different ischemic stroke patient cohorts. The model includes 5 variables, each one with an assigned point value which added resulted in a reliable prediction for the development of seizures over time. The SeLECT score was named after the valuables that make it: Severity of the stroke, Large artery atherosclerosis, Early seizure presentation, Cortical involvement, and MCA Territory. The model had a concordance statistic of 0.77 in validation cohorts with an excellent agreement between the predicted and observed outcomes for seizures at 1 and 5 years.

A similar score aimed to predict PSE in survivors of intracerebral hemorrhage, the CAVE score also follows the acronym of the variables considered: Cortical involvement, Age <65, Volume >10 mL, and Early seizures at presentation <7 days. Each of the variables is assigned 1 point and the summation is associated with a cumulative risk of developing PSE.[10]

Being able to identify the population at increased risk of PSE and the ability to calculate the risk specifically for each patient with stroke, open the doors for targeted patient and family education regarding the presentation, risks, and even prophylaxis for PSE, which so far has not proven efficacious in the prevention of seizures.[11] The identification of increased risk invites us to consider a discussion on seizure precipitants, seizure precautions, driving privileges, or risk of Sudden Unexpected Death in Epilepsy. It also allows us to identify a group of patients who don’t have epilepsy at the moment but will likely develop it within 10 years, which is an ideal model to evaluate the efficacy of antiepileptogenic interventions.[12]

With the available information, we are equipped with the tools needed to identify poststroke patients at high risk of developing PSE. We should consider incorporating seizure presentation, possible complications, and potential treatment to the regular stroke education provided to stroke survivors and their families. The presentation of PSE will likely be in the form of focal seizures, which the patients and their families are more likely to identify as a new stroke than a seizure unless they are educated about both. Asking patients about possible seizure symptoms in a prospective manner at follow-up visits may result in early identification and treatment of PSE. The question of performing prospective electroencephalograms (EEGs) is controversial and seen by many as a waste of resources; however, performing serial EEGs on stroke survivors was able to increase the identification of PSE to 16% of poststroke patients.[4] One more powerful argument comes to mind, cognitive deficits and depression often accompany neurologic conditions such as stroke and epilepsy. The possibility that some of these symptoms could be related to subclinical seizures and therefore could be treated and resolved is enough reason to take PSE seriously as part of stroke management and as a strategy to get our patients to live and function to their highest potential.