Sufentanil, a synthetic opioid that is 5-10 times as potent as fentanyl, has been suggested for use during neurosurgical procedures because it maintains cardiovascular stability and produces hypnosis without the use of additional anesthetic agents. Doses as low as 2.5 micrograms.kg-1 are reported to create deep levels of anesthesia as demonstrated by EEG changes to high-amplitude delta-waves. However, there are no reports concerning the effects of sufentanil on blood flow and metabolism in the human brain. The present study was designed to investigate the influence of high-dose sufentanil-O2 anesthesia on the cerebral circulation, metabolism, and the cerebrovascular response to CO2 in man. METHODS. Nine male and 2 female patients between 41 and 60 years of age who were scheduled for coronary artery bypass surgery were studied. Premedication consisted of flunitrazepam 2 mg orally and piritramide 15 mg and promethazine 50 mg i.m. 1 h before arrival in the induction room. Measurements were performed with the patients awake (I), after sufentanil 10 micrograms.kg-1 as an induction dose followed by 0.15 micrograms.kg-1.min-1 as an infusion with normocapnia (pa CO2 42.1 +/- 2 mmHg) (II), during hypercapnia (pa CO2 53.7 +/- 3.5 mmHg) (III), and during hypocapnia (pa CO2 31.7 +/- 2 mmHg) (IV). Cerebral blood flow (CBF) was measured using the argon wash-in technique. Cerebral venous blood was obtained from a catheter in the superior bulb of the right internal jugular vein. Cerebral metabolic rates of oxygen (CMRO2) glucose (Mgluc) lactate (CMlac) were calculated by multiplying the arterial-cerebral venous oxygen and substrate differences by CBF. The Anaerobic Index was calculated from the equation avD lactate x 100/2 x avD glucose = ANI (%) Cerebral electrical activity was recorded by aperiodic analysis of the EEG (Lifescan). RESULTS AND DISCUSSION. In the EEG sufentanil anesthesia was characterized by a decrease in the number of high-frequency waves and an increase in the number and amplitude of delta-waves, a pattern that did not change throughout the study period. Concomitantly, under normocapnic conditions high-dose sufentanil led to the significant decrease in CBF by 29% accompanied by an 18% increase in cerebral vascular resistance (CVR). CMRO2 decreased by 22% while CMRgluc and CMRlac changed only insignificantly such that the ANI, which represents the percentage of anaerobically metabolized glucose, essentially remained unchanged. Mean perfusion pressure declined by 18% but stayed within the range of autoregulation. Hypoventilation (III) was followed by an 82% increase in CBF as a result of a 55% reduction in CVR, whereas cerebral metabolic parameters did not show important changes when compared to measurement II. Hyperventilation (IV), on the other hand, produced a distinct fall in CBF by 56% to a value that was 21% below the one obtained under normocapnia. This was due to an increase in CVR of the same magnitude. There was a 31% rise in CMRO2, resulting in a decrease in cerebral venous oxygen tension, but in no case did it fall below the critical value of 20 mmHg at which tissue hypoxia becomes severe. Although CMRlac increased and CMRgluc did not significantly change, the ANI remained essentially unchanged, which suggests a predominantly aerobic metabolism. The increase in metabolic activity with sufentanil during hypocapnia might be caused by an alkalosis-induced stimulation of glycolysis. It might also be related to a reduction in the depth of anesthesia, although neither the EEG nor the hemodynamic parameters indicated this. This study shows that the coupling between CBF and metabolism is well maintained and that the cerebrovascular response to CO2 is unimpaired during high-dose sufentanil anesthesia.