Sevoflurane has been shown in animal models to have proconvulsant properties. Nevertheless, with the exception of case reports of the volatile anesthetic triggering seizures in epileptic children, sevoflurane is not generally considered epileptogenic in otherwise healthy patients. Three studies in this issue address changes in electroencephalographic (EEG) activity during sevoflurane anesthesia. In the first article, Kaisti et al.  report two EEG-verified cases of epileptiform activity during a study of the effects of sevoflurane anesthesia on regional cerebral blood flow (rCBF). Eight young healthy volunteers were included in the study. There was no premedication given, and anesthesia was induced via  mask with 7% sevoflurane and 100% oxygen. In addition to positron emission tomography obtained at intervals after increases in end-tidal concentration of anesthetic, there was continuous EEG monitoring.

In the first case, slight clonic movement was noticed 1 h 40 min into anesthesia, when the subject was receiving 2 MAC sevoflurane. End-tidal sevoflurane concentration at the time of positron emission tomography during the seizure was 3.1%. The EEG recording showed a rhythmic epileptiform discharge, a pattern compatible with the clinical seizure type (partial motor seizure). Despite the episode, the subject (who had no history of seizure) awoke from anesthesia and recovered normally. The second patient was the eighth volunteer in the study, who, after 1 h 25 min of sevoflurane anesthesia, had two partial, secondarily generalized EEG discharges, which lasted for 2 and 3 min, respectively. That subject also awoke and recovered normally, and results of magnetic resonance imaging of the brain and clinical neurologic examination were also normal. Although the implications from their findings are unclear, the authors conclude that sevoflurane appears to have epileptogenic potential in healthy patients during burst-suppression anesthesia.

Having noted in a previous study that hyperventilation during sevoflurane nitrous oxide–oxygen mask induction is associated with cardiovascular hyperdynamic response, Yli-Hankala et al.  designed a randomized controlled study to determine whether hyperdynamic response can be explained by EEG findings during sevoflurane mask induction. The team enrolled 30 patients, American Society of Anesthesiologists status I or II, scheduled to undergo gynecologic procedures. All participants were premedicated with oral 5 mg diazepam 1 h before anesthesia. Patients were randomly allocated to one of two groups: spontaneous breathing (SB) and controlled hyperventilation (CH). Baseline heart rate, mean arterial pressure, and oxygen saturation by pulse oximetry (SpO2) were measured; EEG data were collected before and during the study period, then digitized and stored for later analysis by a neurophysiologist blinded to the patients’ ventilation mode.

All patients breathed oxygen via  a clear face mask for 2 min before induction. Anesthesia was induced using the single-breath method with a semiclosed system primed with a fresh gas flow of 10 l/min and a sevoflurane vaporizer. Observers started a stop watch when patients took the first breath from the face mask. The induction period lasted 6 min, at the end of which arterial blood samples were collected for blood gas analysis. Patients were then intubated and the surgery proceeded.

In most of the patients in the SB group, mixed EEG activity, consisting of monophasic slow delta waves with or without spikes, continued to the end of the study. The EEG of seven of the SB group patients and all patients in the CH group showed epileptiform EEG activity—periods of polyspikes or rhythmic polyspikes. Heart rate increased 54% more than baseline values in the CH group 4 min after induction. The epileptiform EEG patterns elicited by sevoflurane mask induction may be explained by the speed of the anesthetic induction or the ventilation mode itself.

Constant et al.  also used continuous EEG, in addition to heart rate and finger blood pressure, to evaluate EEG tracings and autonomic cardiovascular activity after induction with either sevoflurane or halothane in children aged 2–12 yr of age scheduled for elective tonsillectomy. After premedication with midazolam, patients were randomly assigned to one of three induction techniques: rapid induction with 7% sevoflurane in 100% O2; incremental induction with 2, 4, 6, and 7% sevoflurane every five breaths in 100% O2; or incremental induction with 1, 2, 3, and 3.5% halothane every five breaths in a 50:50 mixture of oxygen and nitrous oxide.

An additional group of 10 patients was later enrolled after completion of the first study in an open-label, nonrandomized arm. Patients received 7% sevoflurane in a 50:50 mixture of oxygen and nitrous oxide. This was deemed necessary because nitrous oxide was used in the original study's halothane group but was omitted in the sevoflurane groups. The same monitoring of EEG and hemodynamic data was performed.

In contrast to the observations made in the Kaisti et al.  and Yli-Hankala et al.  studies, researchers found no seizure-like activity in the 45 EEG tracings obtained. Induction of anesthesia was associated in all four groups with an increase in TSP and a shift toward the low-frequency bands. Sevoflurane induced greater withdrawal of parasympathetic activity than halothane, and transient relative increases in sympathetic vascular tone at loss of eyelash reflex.