WS1.11. EEG in the ICU: Status Epilepticus and the Ictal-Interictal Continuum

Abstract

Epileptic seizures and nonconvulsive status (NCSE) are common in the critically ill, but non- epileptic movements and posturing are even more common. EEG can provide clarification when seizures are suspected. Different EEG patterns form a continuum of non-seizure to seizure activity, and over- interpretation of these patterns may lead to inappropriate advice and treatment. Recurrent seizures are easiest to recognise, with rhythmic discharges having spatiotemporal evolution in frequency (>4Hz), morphology and distribution. Periodic patterns are the most difficult. Generalised periodic discharges (GPDs) may or may not be ictal, and the specific EEG findings combined with the clinical context determines the approach. Severe myoclonic encephalopathy may occur after cardiac arrest, and the EEG may show burst suppression, unreactive GPDs separated by isoelectric intervals, ± electrographic seizures. The prognosis is extremely poor irrespective of treatment. Lateralised periodic discharges (i.e. PLEDs) may be an interictal finding in epilepsy and after any acute cerebral insult such as stroke or herpes encephalitis, and usually indicates a predisposition to seizures rather than ongoing seizure activity. Generalised triphasic waves are mainly seen with metabolic and toxic encephalopathies and are usually not an ictal finding. Whilst they can be sharply contoured, they are recognised by their other characteristics including morphology, state-dependent reactivity, and lag in phase and/or morphology changes from front to back. Triphasic waves (and almost all periodic patterns) are suppressed by benzodiazepines, along with consciousness and respiration. Such EEG suppression, in the absence of clinical improvement, does not diagnose NCSE. Ambiguous patterns can also be seen, with stimulus-induced rhythmic, periodic or evolving discharges. Whilst non-convulsive seizures and NCSE may be detected in comatose patients, prognosis is largely determined by the underlying aetiology and its complications. Unresolved questions include what particular EEG patterns should consistently prompt aggressive treatment. It is not always possible to be certain that an EEG pattern is ictal or non- ictal, and in the correct clinical context (especially patients presenting with seizures or with a history of epilepsy) a treatment trial may be indicated to see if temporally associated clinical improvement occurs. EEG interpretation needs to be made within the clinical context by an expert reporter.