Seizures Worsen Stroke Outcome: New Evidence From a Large Sample

Commentary

Influence of Seizures on Stroke Outcomes: A Large Multicenter Study.

Huang CW, Saposnik G, Fang J, Steven DA, Burneo JG. Neurology 2014;82:768–776.

OBJECTIVE: We compared clinical characteristics of seizures at ischemic stroke presentation (SSP) to seizures during hospitalization post ischemic stroke (SDH), and their impacts on stroke outcome, using the Registry of the Canadian Stroke Network (RCSN) database. METHODS: This cohort study included consecutive patients from the RCSN who had an acute ischemic stroke between July 2003 and March 2008. Outcome measures included morbidity, mortality, length of hospital stay, and discharge disposition. Clinical variables for either SSP or SDH were investigated and the stroke outcome was stratified by stroke severity. RESULTS: The study included 10,261 patients with ischemic strokes: 157 patients (1.53%) had SSP and 208 patients (2.03%) had SDH. Compared to stroke patients without seizures, patients with SSP and SDH were younger, had more severe strokes (p < 0.001), a higher admission rate to the intensive care unit (p < 0.001), higher morbidity, and higher mortality (p < 0.05). SSP was associated with female sex and less limb weakness, while SDH was associated with pneumonia and the presence of hemineglect. Importantly, patients with less severe strokes had higher morbidity and mortality (p < 0.005) if SDH occurred. Variables predicting overall mortality were SDH, older age, higher Charlson-Deyo index, more severe strokes, and nonalert status on arrival (all p < 0.001). CONCLUSIONS: SSP and SDH have different characteristics. SDH indicates a poorer prognosis in patients. Increased awareness of SSP and efforts to prevent SDH may be important in improving outcomes following clinical stroke care.

Cerebrovascular disease is the most common cause of new-onset epilepsy in older adults, accounting for more than one-third of all cases (1). The occurrence of seizures is known to affect stroke outcome. In a cohort of 5,027 stroke patients, seizures occurred in 2.7% of the cases and predicted increased mortality, longer hospitalization, and more disability (2). Seizures can occur early or late after an ischemic stroke. While approximately 3 to 5 percent of strokes are associated with a late seizure within the first year, up to 66% of these patients will develop epilepsy (4). In a population-based study in Rochester, MN, 535 consecutive persons were followed after their first cerebral infarction, and 33 (6%) had their first seizure within the first week (3). Of these, 78% had seizures in the first day after stroke. Risk for early seizures increased with anterior hemisphere location of infarct. New onset late seizures (past the first post-stroke week) occurred in 5%, and their risk was increased by early seizures and stroke recurrence, which are also risk factors for the development of epilepsy.

Stroke has also been shown to increase the risk of status epilepticus. In one series, up to 25% of patients with status epilepticus presented with stroke (5). There are no guidelines about prophylactic treatment with antiseizure medications, or even treating after the first seizure, since the effect of a second seizure in an untreated stroke patient has not been proven to affect the overall outcome (4).

Early and late seizures may differ both in terms of their underlying mechanisms and risk of epilepsy. Unlike early seizures that may result from biochemical imbalances, late seizures may result from gliotic scarring. In acute ischemia, glutamate-mediated excitotoxicity leads to an overload of intracellular calcium, which activates a number of intracellular enzymes and immediate early genes, eventually culminating in neuronal death. In surviving neurons, glutamate can induce epileptiform discharges (6). Thus, seizures in acute ischemia can increase the metabolic demand in the hypoxic tissue, potentially leading to increased infarct size (2). In contrast, late seizures may result from gliotic scarring with associated changes in membrane properties, deafferentation, selective neuronal loss, and collateral sprouting—all leading to hyperexcitability and increased synchronization of neuronal activity (7).

In an attempt to test the hypothesis whether the pathophysiology of seizures after ischemic stroke is determined by the stage of the stroke, Huang et al. used the Registry of the Canadian Stroke Network, which collects data prospectively from multiple centers, to investigate the clinical features and outcomes of seizures occurring at stroke presentation (SSP) and seizures occurring after stroke but during hospitalization (SDH) (8). Their cohort consisted of 10,261 patients with ischemic stroke. The authors found that seizures were more likely to occur in individuals with more severe strokes, of younger age, and were associated with higher rate of admission to the intensive care unit, higher morbidity, and higher mortality. The rate of seizure occurrence was consistent with previous studies (9): SSP occurred in 157 patients (1.53%) and correlated with female sex and less limb weakness. SDH occurred in 208 patients (2.03%) and was associated with pneumonia and hemineglect. SDH correlated with a mortality rate of 30.3% at 1 month and 47.6% at 1 year, compared with 13.3% and 24.1%, respectively, for patients without seizures. In addition, SDH was associated with higher morbidity and mortality even in patients with less severe strokes. SSP was an important risk factor for SDH, constituting an odds ratio of 15.1 for SDH occurrence.

This study has a number of strengths, including the large sample size, studying only ischemic stroke, and the prospective collection of important data, including laboratory and radiologic investigations, clinical assessment of stroke and seizure variables, and follow up. In addition, the differences noted by the authors between SSP and SDH are an important first attempt at shedding light on possibly different seizure pathophysiology in these two presentations. The authors found that SSP necessitated fewer interventions, including thrombolytic therapy and occupational therapy, possibly because SSP occurred with more severe strokes. That said, SDH correlated with higher rates of depression and infections, including pneumonia and urinary tract infection, as well as decubitus ulcer, suggesting that these factors may play an additional role in lowering the seizure threshold.

This study confirms known facts about the low incidence of acute seizures after ischemic stroke and the negative influence of seizures on stroke outcomes. While the authors did not directly address the important question of seizure prevention after stroke, they highlighted risk factors for seizures in a large population. The low incidence of seizures argues that prophylactic use of antiseizure medications is not warranted, and knowledge of seizure risk factors helps the practitioners individualize the decision to start an antiseizure medication after the first seizure. The fact that SSP is an important predictor of SDH suggests use of medications for secondary prevention of seizures after SSP. Other factors that need to be taken into account include stroke severity, location (e.g., cortical involvement), the presence of infections, depression, and even hemineglect. The paucity of evidence for the use of antiseizure medications after ischemic stroke should be addressed by a future study that balances the potential benefit of preventing stroke-related seizures by using long-term antiseizure medications against the negative effects of some medications, such as phenytoin, on neurological recovery (10). Another important question that warrants investigation is related to the duration of treating stroke-related seizures.