As is typical in seizure diagnoses, the most important data derive from the history, which usually must be provided by observers as well as the patient. Topics should include:
The neurologic examination typically reflects the location and severity of the stroke, although an examination performed shortly after a seizure usually shows a more severe deficit. If a large stroke or other structural lesion is present, such a postictal or Todd’s paralysis may persist for hours or days.
Neuroimaging abnormalities correspond primarily to the stroke itself, although after seizures (and particularly after status epilepticus), a variety of imaging changes sometimes persist for days or weeks.87 These usually affect cortex as well as white matter, and include
These changes may be more confusing than enlightening in trying to determine whether seizures have occurred, but serial imaging correlated with the changing neurologic exam can be crucial to understanding the clinical process.
Perfusion imaging with MRI or single photon emission computerized tomography (SPECT) is also abnormal after stroke, showing hypoperfusion sometimes followed, after a week or more, by surrounding hyperperfusion. Perfusion images in status epilepticus, by contrast, show an increase ictally, usually a decrease postictally for minutes or longer, and then normalization (or return to the abnormal baseline in the case of a stroke). Similar changes can occur even after single seizures, but ictal SPECT in non–status epilepticus cases is seldom available outside the epilepsy monitoring unit.
Despite controversies about predictive value, the electroencephalogram (EEG) remains the fundamental investigation for patients with seizures. The main EEG manifestation after stroke is focal slowing. The development of a seizure focus may be accompanied by the appearance of epileptiform discharges such as sharp waves, spikes, or periodic lateralized epileptiform discharges (PLEDs). Studies suggest that interictal epileptiform discharges are less commonly seen in older than in younger patients with established epilepsy,88 but similar data are not available for acute seizures. On the other hand, sharp waves may occur even in studies of patients who do not have seizures. Well-defined spike foci may have higher predictive value than sharp waves, although detailed evidence is lacking. Most studies, however, suggest that PLEDs are a strong predictor of acute seizures.57,64,89,90 Focal slowing alone may also be somewhat predictive.47,86,91 EEG interpretation in the elderly must take into account that normal variants, particularly sharply contoured, alpha-frequency “wickets” in the temporal regions, are not indicators of a seizure tendency.
EEG manifestations of nonconvulsive status epilepticus require particular attention. The easiest pattern to diagnose, electrographically as well as clinically, is that of repeated individual seizures with incomplete recovery between episodes. More difficult are those that show continuous rhythmic activity, at times without well-defined spikes or sharp waves.82,83 This must be distinguished from other causes of more or less rhythmic slowing, especially those associated with metabolic disturbances. The triphasic wave pattern, classically and most commonly seen with hepatic encephalopathy but also associated with uremia, anoxia, and other conditions, is a particular source of confusion. A trial of intravenous benzodiazepines can sometimes clarify the situation if the EEG and patient improve, but one must keep in mind that the triphasic pattern usually disappears in sleep.
Long-term video-EEG monitoring is probably underused in elderly patients in general88 and in stroke patients in particular. This should be considered when any patient has relatively frequent events that could represent seizures or that have been treated as seizures with inadequate response. Concurrent monitoring of ECG and other physiologic parameters, such as respiratory effort and oxygen saturation, may sometimes suggest alternative diagnoses.
Laboratory evaluation is aimed at ruling out metabolic factors that could predispose to seizures.81 Serum chemistries, particularly sodium, calcium, magnesium, glucose, and renal indices, should be measured. Levels of potentially offending drugs, such as theophylline or the normeperidine metabolite of meperidine, can at times be obtained rapidly enough to be useful. Blood and urine toxic screens can be helpful in assessing the possibility of illicit drug use, particularly in younger patients with stroke and seizures. Cocaine is of particular importance.
Lumbar puncture is rarely a consideration in stroke patients, except when subarachnoid hemorrhage is considered and CT is negative for blood, or when the situation suggests that stroke could have resulted from an infectious vasculitis secondary to meningitis. In immunocompromised patients, for example, central nervous system aspergillosis is frequently accompanied by hemorrhage, as a result of fungal invasion through the vessel wall.
Reviewed and revised April 2004 by Steven C. Schachter, MD, epilepsy.com Editorial Board.
© 2014 Epilepsy.com. All rights reserved.