Units

Contact

---

Citation

Type

Conference Name

Publication Date

Brief description/objective

Objectives: The significance of an early electroencephalogram (EEG) for the prediction of poststroke seizures is unknown and has not been systematically evaluated yet.
Methods: Included were patients with first-ever ischemic stroke who received an early poststroke EEG. Cases with other potentially epileptogenic conditions were excluded. Follow-up of patients was performed by telephone or at our neurologic clinic at least 2 years after stroke.
Results: We have screened 541 consecutive patients and included 335 in final analysis. A total of 46 (14%) patients developed epileptic seizures, whereas early-onset (<= 7 days) and late-onset (> 7 days) seizures occurred in 24 (7%) and 28 (8%) patients respectively. Cases with seizures were more likely to have a cortical infarction (odds ratio [OR]=4.3, p<0.001), a higher stroke severity (NIHSS>4, OR=2.8, p=0.003) and a macroangiopathic etiology of stroke (OR=2.2, p=0.05). There were no between group differences in age, gender, stroke risk factors or arterial territory, however patients with seizures had a significantly higher mortality (22% vs. 8%, p=0.01). An abnormal EEG predicted seizures with an OR of 4.4 (p<0.001). A moderate to severe focal slowing (OR=4.5, p<0.001) and (frontal) intermittent rhythmic delta activity (F/IRDA, OR=5.4, p=0.02) were significant EEG signs, whereas diffuse slowing, epileptiform discharges and periodic lateralized epileptiform discharges (PLEDs) were of no significance. In a multivariate logistic regression model an abnormal EEG (adjusted OR=3.4, p=0.004) and cortical stroke location (adjusted OR=3.3, p=0.005) were the only significant individual predictors of seizures.
Conclusion: An abnormal early EEG might be the most significant independent predictor of seizures after stroke. Focal slowing and F/IRDA are individual EEG signs associated with poststroke seizures.