Present package labeling for sevoflurane recommends the use of fresh g
as flow rates of 2 L/min or more when delivering anesthesia with sevof
lurane. This recommendation resulted from a concern about the potentia
l nephrotoxicity of a degradation product of sevoflurane, ''Compound A
,'' produced by the action of carbon dioxide absorbents on sevoflurane
. To assess the adequacy of this recommendation, we compared the nephr
otoxicity of 8 h of 1.25 minimum alveolar anesthetic concentration (MA
C) sevoflurane (n = 10) versus desflurane (n = 9) in fluid-restricted
(i.e., nothing by mouth overnight) volunteers when the anesthetic was
given in a standard circle absorber anesthetic system at 2 L/min. Subj
ects were tested for markers of renal injury (urinary albumin, glucose
, alpha-glutathione-S-transferase [GST], and pi-GST; and serum creatin
ine and blood urea nitrogen [BUN]) before and 1, 2, 3, and/or 5-7 clay
s after anesthesia. Desflurane did not produce renal injury. Rebreathi
ng of sevoflurane produced average inspired concentrations of Compound
A of 41 +/- 3 ppm (mean +/- SD). Sevoflurane was associated with tran
sient injury to: 1) the glomerulus, as revealed by postanesthetic albu
minuria; 2) the proximal tubule, as revealed by postanesthetic glucosu
ria and increased urinary alpha-GST; and 3) the distal tubule, as reve
aled by postanesthetic increased urinary pi-GST. These effects varied
greatly (e.g., on postanesthesia Day 3, the 24-h albumin excretion was
<0.03 g (normal) for one volunteer; 0.03-1 g for five others; 1-2 g f
or two others; 2.1 g for one volunteer; and 4.4 g for another voluntee
r). Neither anesthetic affected serum creatinine or BUN, nor changed t
he ability of the kidney to concentrate urine in response to vasopress
in, 5 U/70 kg subcutaneously (i.e., these measures failed to reveal th
e injury produced). In addition, sevoflurane, but not desflurane, caus
ed small postanesthetic increases in serum alanine aminotransferase (A
LT), suggesting mild, transient hepatic injury.