The Safety of Sevoflurane Has Not Been Adequately Established

To the Editor: Sevoflurane was recently approved by the Food and Drug Administration (FDA) for use in general anesthesia. Of concern to the FDA was the in vivo and in vitro degradation of sevoflurane. Specifically, carbon dioxide absorbents degrade sevoflurane to Compound A [CH2 F bond O bond C(double bond CF2)(CF3)], a compound that can cause corticomedullary renal necrosis in rats. The threshold for renal injury is 25-114 ppm for 3-12 h of exposure [1-3]. Low-flow systems used to deliver sevoflurane to patients can produce concentrations of Compound A of 20-40 ppm [4-7]. Thus, the Compound A concentrations occurring clinically differ little from concentrations that injure rats. The similarity of the clinically achieved concentrations to those that injure rats is of concern [8]. Studies in humans do not adequately address the issue of the safety of Compound A. Only low inflow rates containing sevoflurane can provide Compound A in respired gases. Thus, the extensive Japanese experience with sevoflurane does not furnish evidence of innocuousness because inflow rates in that country normally exceed 3 L/min [9]. Similarly, the limited studies from other countries commonly apply inflow rates exceeding 4 L/min [10]. Human studies also may fail to address the issue of the safety of Compound A because they apply crude measures of renal function such as creatinine and blood urea nitrogen levels [6,11-16] rather than sensitive measures of renal integrity. The enormous functional reserve of the kidney would prevent such tests from revealing injury unless a large portion of the kidney were compromised. Of greater importance, serum creatinine concentrations provide evidence for glomerular not tubular integrity. Compound A injures renal tubules. Finally, one may question whether Compound A, a halogenated vinyl ether, equally affects humans and rats. Several halogenated alkenes produce a corticomedullary lesion analogous to that produced by Compound A. Many alkenes injure by conversion of their glutathione conjugates to cysteine S-conjugates and subsequent metabolism by renal cysteine conjugate beta-lyase to nephrotoxic halothionoacetyl halides [17-20]. Because the enzymes mediating these reactions have a far greater activity in rats than in humans [21,22], it may be argued that humans are at a commensurately lower risk of injury from Compound A. Controversy exists on this point. A recent report supports the argument [23], finding that blockade of this pathway decreases diuresis and proteinurea induced by intraperitoneal injection of Compound A. In conflict, we find that blockade of these enzymes increases histologically observed injury from inhalation of Compound A [24]. Approval of sevoflurane by the FDA followed the finding by the Anesthesia and Life Support Drugs Advisory Committee that the degradation of sevoflurane to Compound A poses a potential, presently unknown, risk of renal injury to patients. The Committee recommended that the package labeling should include a warning to this effect. The Committee also recommended use of inflow rates of 2 L/min or more. One approach to safety applies a limit at least 10-fold less than the concentration known to cause injury [8]. An inflow rate of 2 L/min does not provide this protection, particularly for patients receiving high concentrations of sevoflurane for prolonged periods of time. In summary, studies to date do not exclude the possibility that degradation of sevoflurane to Compound A can produce renal injury, particularly a subtle injury, in humans. Studies that used lower inflow rates have not applied measurements that would test for a subtle injury to renal tubules. Studies that use high inflow rates preclude Compound A from reaching the patient and therefore cannot be used to document the safety of Compound A. The potential for Compound A-induced renal injury at lower inflow rates remains a cause for concern. Conflict of interest statement: Dr. Eger is a paid consultant to Ohmeda (Liberty Corner, NJ), Pharmaceutical Products Division. Drs. Eger and Martin are recipients of grants-in-aid from Ohmeda for the conduct of research. Dr. Tinker is a co-inventor of an analog of sevoflurane. He is engaged in efforts to develop this new anesthetic and methodologies to inhibit Compound A formation from sevoflurane. E. I Eger II, MD Department of Anesthesia University of California, San Francisco San Francisco, CA 94143-0464 J. L. Martin, MD Department of Anesthesiology Johns Hopkins University Baltimore, MD 21287 J. H.