Sevoflurane, a new inhalational anesthetic agent has been shown to produce degradation products upon interaction with CO2 absorbants. Quantification of these sevoflurane degradation products during low-flow or closed circuit anesthesia in patients has not been well evaluated. The production of sevoflurane degradation products was evaluated using a low-flow anesthetic technique in patients receiving sevoflurane anesthesia in excess of 3 h. Sevoflurane anesthesia was administered to 16 patients using a circle absorption system with O2 flow of 500 ml/min and average N2O flow of 273 ml/min. Preoperative and postoperative hepatic and renal function studies were performed. Gas samples were obtained from the inhalation and exhalation limbs of the anesthetic circuit for degradation product analysis and analyzed by gas chromatography/mass spectrometry for four degradation products. The first eight patients received sevoflurane anesthesia using soda lime, and the following eight patients received anesthesia using baralyme as the CO2 absorbant. CO2 absorbant temperatures were measured during anesthesia. Of the degradation products analyzed, only one compound [fluoromethyl-2, 2-difluoro-1-(trifluoromethyl) vinyl ether], designated compound A, was detectable. Concentrations of compound A increased during the first 4 h of anesthesia with soda lime and baralyme and declined between 4 and 5 h when baralyme was used. Mean maximum inhalation concentration of compound A using baralyme was 20.28 +/- 8.6 ppm (mean +/- SEM) compared to 8.16 +/- 2.67 ppm obtained with soda lime, a difference that did not reach statistical significance. A single patient achieved a maximal concentration of 60.78 ppm during low-flow anesthesia with baralyme. Exhalation concentrations of compound A were less than inhalation concentrations, suggesting patient uptake. Peak CO2 absorbant temperatures were greater with baralyme (46.4 +/- 1.31-degrees-C) than soda lime (37.8 +/- 1.14-degrees-C) (P < .001). No abnormal changes in hepatic or renal function testing were detectable up to 48 h after anesthesia. Sevoflurane when used in a low-flow anesthetic circuit for 3-5 h duration generally produced low levels of a single degradation product, though in one case a higher concentration of compound A was produced. Although the concentrations of compound A obtained in this study are well below those reported to cause toxicity in animals, further studies will be needed to evaluate potential variability in compound A production. No patient in our study had clinical or laboratory evidence of organ toxicity following low-flow sevoflurane anesthesia.
