Electrographic signatures of postanoxic brain injury

After a successful resuscitation from cardiac arrest, most patients remain comatose as a result of postanoxic encephalopathy. More than half of them never regain consciousness, and treatment options to improve their outcome are limited. The aim of the research described in this dissertation was to validate and improve the value of continuous electroencephalography (EEG) for outcome prediction and treatment of postanoxic brain injury. In a prospective cohort study of 850 patients, we confirmed that EEG reaches its maximum value for the prediction of outcome within the first 24 hours after cardiac arrest. The added value of continuous EEG monitoring beyond this period was limited. Generalized suppression (all EEG activity <10 µV) or synchronous patterns with more than 50% suppression reliably predicted a poor outcome at 12h after cardiac arrest or later. Continuous background activity within 12h from cardiac arrest was a strong predictor of good outcome. To make the assessment of postanoxic EEG less time-consuming and more objective, we introduced straightforward quantitative EEG features, based on key aspects of visual assessment for the prediction of outcome. Our measures for background continuity and amplitude fluctuation were at least as sensitive for the prediction of good outcome as visual assessment, at equal reliability. In the subgroup of patients with electrographic seizure activity, we showed that a lack of background continuity of the EEG precludes recovery. During the first 24 hours after cardiac arrest, the most valuable period for the prediction of outcome, patients are usually treated with sedative medication. We showed quantitatively that propofol, a commonly applied sedative drug, changes the postanoxic EEG, but does not affect its reliability for the prediction of outcome. A better understanding of mechanisms that underlie postanoxic EEG patterns could validate associations between EEG and outcome, and offer opportunities for new treatment strategies. By using a computational model, we showed that pathophysiological changes at the synaptic level explain the most commonly observed EEG patterns after cardiac arrest and their evolution. Finally, we present the study protocol of the ongoing, randomized TELSTAR trial on the treatment of electrographic status epilepticus after cardiac arrest.