Free
Correspondence  |   September 2003
Safety of Low-flow Sevoflurane Anesthesia in Patients with Chronically Impaired Renal Function is not Proven: In Reply
Author Affiliations & Notes
  • Evan D. Kharasch, M.D., Ph.D.
    *
  • *University of Washington, Seattle, Washington.
Article Information
Correspondence
Correspondence   |   September 2003
Safety of Low-flow Sevoflurane Anesthesia in Patients with Chronically Impaired Renal Function is not Proven: In Reply
Anesthesiology 9 2003, Vol.99, 752-754. doi:
Anesthesiology 9 2003, Vol.99, 752-754. doi:
In Reply:—
We appreciate the interest of Drs. Saidman and Eger in the safety of sevoflurane and the protection of human research participants. Although we disagree with their contentions, we completely share their safety concerns. Indeed, our own concern for patient safety is why we performed the investigation. We assessed the renal effects of low-flow (≤ l/min) sevoflurane in patients at greatest risk for postoperative renal dysfunction: those with preexisting renal insufficiency. Even in such susceptible patients, the renal effects of low-flow sevoflurane and isoflurane were the same. 1 Our conclusions were specific: “Low-flow sevoflurane is as safe as low-flow isoflurane and does not alter renal function in patients with preexisting renal disease.” These results amplify previous studies in patients with renal insufficiency, conducted at higher flow rates, which showed no significant differences in the renal effects of sevoflurane and other volatile anesthetics. 2–6 
This investigation helps to resolve any outstanding questions about the renal effects of sevoflurane. These have concerned sevoflurane defluorination, patients with renal insufficiency, low flows and compound A formation, and low-flows in renal insufficiency patients. What has emerged from prospective studies and from postapproval pharmacovigilance is a remarkably consistent picture. Postoperative sevoflurane renal effects are not different from those of other anesthetics. After more than 120 million sevoflurane anesthetics given, there is not a single case report of sevoflurane-related renal dysfunction. Considering together all of the studies published to date in patients or volunteers, and even using proteinuria as a so-called “sensitive” (albeit unvalidated and experimental) marker of renal dysfunction, there is no difference between the renal effects of low-flow sevoflurane and other anesthetics (Fig. 1).
Fig. 1. Effect of anesthesia on urine protein excretion. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Low-flow (< 2 l/min) sevoflurane data are from surgical patients (n = 120), 15–20 surgical patients with chronic renal insufficiency (n = 56), 1 and normal volunteers (n = 71). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 Comparator data are from low-flow isoflurane in patients with normal renal function (n = 83), 15,17–19 patients with renal insufficiency (n = 53), 1 and volunteers (n = 4) 8,9; high-flow sevoflurane patients (n = 40) 15,16,19; low-flow desflurane patients (n = 18) 20; and propofol patients (n = 10). 20 Box plots show the median, mean (dashed lines  ), 25th and 75th percentiles (box boundaries  ), 10th and 90th percentiles (whiskers  ), and outliers outside the 10th and 90th percentiles. The reference range based on data from healthy nonsurgical subjects is shown by the dotted line. There were no significant differences between the groups (P  = 0.25, Mann–Whitney Rank Sum test).
Fig. 1. Effect of anesthesia on urine protein excretion. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Low-flow (< 2 l/min) sevoflurane data are from surgical patients (n = 120), 15–20surgical patients with chronic renal insufficiency (n = 56), 1and normal volunteers (n = 71). 7–10,21Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10Comparator data are from low-flow isoflurane in patients with normal renal function (n = 83), 15,17–19patients with renal insufficiency (n = 53), 1and volunteers (n = 4) 8,9; high-flow sevoflurane patients (n = 40) 15,16,19; low-flow desflurane patients (n = 18) 20; and propofol patients (n = 10). 20Box plots show the median, mean (dashed lines 
	), 25th and 75th percentiles (box boundaries 
	), 10th and 90th percentiles (whiskers 
	), and outliers outside the 10th and 90th percentiles. The reference range based on data from healthy nonsurgical subjects is shown by the dotted line. There were no significant differences between the groups (P 
	= 0.25, Mann–Whitney Rank Sum test).
Fig. 1. Effect of anesthesia on urine protein excretion. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Low-flow (< 2 l/min) sevoflurane data are from surgical patients (n = 120), 15–20 surgical patients with chronic renal insufficiency (n = 56), 1 and normal volunteers (n = 71). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 Comparator data are from low-flow isoflurane in patients with normal renal function (n = 83), 15,17–19 patients with renal insufficiency (n = 53), 1 and volunteers (n = 4) 8,9; high-flow sevoflurane patients (n = 40) 15,16,19; low-flow desflurane patients (n = 18) 20; and propofol patients (n = 10). 20 Box plots show the median, mean (dashed lines  ), 25th and 75th percentiles (box boundaries  ), 10th and 90th percentiles (whiskers  ), and outliers outside the 10th and 90th percentiles. The reference range based on data from healthy nonsurgical subjects is shown by the dotted line. There were no significant differences between the groups (P  = 0.25, Mann–Whitney Rank Sum test).
×
Drs. Saidman and Eger assert that the compound A exposure in our investigation was “too small.” The patients received the compound A exposure they did because the average low-flow duration was 3.2 h, and compound A concentrations typically average 10–15 ppm. Clinical research is captive to the patient population at hand. Nevertheless, we used all available means to maximize compound A concentrations. The only way to increase compound A exposures even more would have been to prolong the anesthetic far beyond that needed for surgery, purely for research purposes. We would argue against the ethics of this approach. The study conditions reflected typical clinical care. This is a meaningful test of the hypothesis. Under relevant clinical conditions, low-flow sevoflurane and isoflurane effects were the same.
Our colleagues also assert that the compound A exposure was “too small” because it was below the “threshold for renal injury” of 150–200 ppm-hr. We believe that this premise is false. The results of Eger et al.  7 in volunteers, on which the purported human “threshold” is based, have never been replicated despite all best efforts. 8–10 Results in patients similarly show no threshold (published studies, and as summarized in Fig. 2).
Fig. 2. Relationship between protein excretion and compound A exposure (area under the curve of inspired compound A concentration vs.  time) during low-flow sevoflurane anesthesia. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Data are from surgical patients (n = 94), 15–20 and normal volunteers (n = 68). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 
Fig. 2. Relationship between protein excretion and compound A exposure (area under the curve of inspired compound A concentration vs. 
	time) during low-flow sevoflurane anesthesia. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Data are from surgical patients (n = 94), 15–20and normal volunteers (n = 68). 7–10,21Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10
Fig. 2. Relationship between protein excretion and compound A exposure (area under the curve of inspired compound A concentration vs.  time) during low-flow sevoflurane anesthesia. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Data are from surgical patients (n = 94), 15–20 and normal volunteers (n = 68). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 
×
Drs. Saidman and Eger also complain that the data were “skewed.” From a statistical perspective, this is correct—the data were not normally distributed. However, the “skewness” occurred in both anesthetic groups and in this case indicates no inadequacy in study design or conduct. It simply represents a pattern of interindividual variability that is often seen in clinical studies. Nature does not require biologic variability to follow a Gaussian distribution. This variability obscured no “modest nephrotoxicity,” and, most specifically, it did not obscure the result that creatinine concentrations (the definitive standard of renal function) were increased in 12% of sevoflurane patients and 14% of patients anesthetized with isoflurane (despite the absence of compound A or meaningful increases in plasma fluoride). Clearly, it is neither the choice of anesthetic nor the flow rate at which it is delivered that determines postoperative renal function.
The hypothesis that sevoflurane has adverse clinical renal effects has been tested and rejected. Nevertheless, Drs. Saidman and Eger continue to suggest that a “controversy” exists (termed a “crusade” by one editorialist 11). It does not.
Let us provide some additional perspective: The anesthesia community knew of methoxyflurane nephrotoxicity within a year after the drug was introduced. 12 The problem of “halothane hepatitis” became abundantly clear after a few million halothane anesthetics. 13 It has now been nearly a decade and more than 120 million patients since the introduction of sevoflurane. There is still no evidence that sevoflurane causes nephrotoxicity.
We would next like to address the use of sevoflurane at low-flow durations beyond those recommended by the Food and Drug Administration, which Drs. Saidman and Eger question. The study was performed specifically in response to the Food and Drug Administration, which requested further evaluation of sevoflurane safety under low-flow conditions and in patients with renal insufficiency (beyond studies submitted for initial regulatory approval). Furthermore, our study in renal insufficiency patients was performed only after several well-controlled, randomized, monitored, multicenter studies of low-flow sevoflurane in healthy volunteers and in patients with normal  kidney function showed no difference in renal effects compared to other volatile anesthetics. The phase IV protocol in our study, including design and safety aspects, was reviewed and specifically approved by the Food and Drug Administration. It was also independently monitored, and an interim analysis was performed to ensure patient safety.
Drs. Eger and Saidman also question the informed consent used in our study. They ask what information was provided to the U. S. Institutional Review Boards and to the patients. The applications to the respective Review Boards described the published literature on sevoflurane renal effects and contained the sevoflurane and isoflurane package inserts. Good Clinical Practice requires that the informed consent process, and the consent document that patients read and sign, plainly describe the relevant and reasonably foreseeable  risks to subjects. The consent document should not itemize the results of individual and controversial scientific articles, debate the literature, or replicate verbatim the package insert. In our study, the relevant and reasonably foreseeable risks were plainly described in the informed consent documents, which were signed by the research subjects. All of the multiple institutional Review Boards that approved the investigation also approved the informed consent document.
Many years ago, a concern was raised about low-flows and the degradation of another anesthetic, halothane. An author responded: “No new evidence indicates that a risk accrues to the degradation of halothane. When previous experiments have been thorough and the nature of the compound and its metabolites investigated and no toxicity found, isn't this a reasonable working definition of nontoxicity? [This is] the problem of discovering nonexistence. Obviously, when what you're searching for doesn't exist, you'll have trouble finding it even with an infinite number of experiments. Although halothane (or enflurane or diazepam or Innovar) may not be toxic, you cannot construct a study that will conclusively document nontoxicity. Long ago my father warned me that I could not disprove the existence of dragons.” That eloquent defender of low-flow halothane was Dr. Eger. 14 
Now, substitute sevoflurane for halothane in the above paragraph. Previous sevoflurane experiments have been very thorough. Over 120 million patients have safely received sevoflurane at both high and low gas-flow rates. There is still no evidence of sevoflurane renal toxicity in surgical patients or in volunteers (other than that reported by Dr. Eger). The nature of sevoflurane and its metabolites and degradation products are more thoroughly understood than any other volatile anesthetic in history. Isn't this a reasonable working definition of nontoxicity? Science and medicine still cannot disprove the existence of dragons.
It is time to discard the unfounded concerns of sevoflurane nephrotoxicity.
References
Conzen PF, Kharasch ED, Czerner SF, Artru AA, Reichle FM, Michalowski P, Rooke GA, Weiss BM, Ebert TJ: Low-flow sevoflurane compared with low-flow isoflurane anesthesia in patients with stable renal insufficiency. A nesthesiology 2002; 97: 578–84Conzen, PF Kharasch, ED Czerner, SF Artru, AA Reichle, FM Michalowski, P Rooke, GA Weiss, BM Ebert, TJ
Conzen PF, Nuscheler M, Melotte A, Verhaegen M, Leupolt T, Van Aken H, Peter K: Renal function and serum fluoride concentrations in patients with stable renal insufficiency after anesthesia with sevoflurane or enflurane. Anesth Analg 1995; 81: 569–75Conzen, PF Nuscheler, M Melotte, A Verhaegen, M Leupolt, T Van Aken, H Peter, K
Tsukamoto N, Hirabayashi Y, Shimizu R, Mitsuhata H: The effects of sevoflurane and isoflurane anesthesia on renal tubular function in patients with moderately impaired renal function. Anesth Analg 1996; 82: 909–13Tsukamoto, N Hirabayashi, Y Shimizu, R Mitsuhata, H
Nishiyama T, Aibiki M, Hanaoka K: Inorganic fluoride kinetics and renal tubular function after sevoflurane anesthesia in chronic renal failure patients receiving hemodialysis. Anesth Analg 1996; 83: 574–7Nishiyama, T Aibiki, M Hanaoka, K
Mazze RI, Callan CM, Galvez ST, Delgado-Herrera L, Mayer DB: The effects of sevoflurane on serum creatinine and blood urea nitrogen concentrations: A retrospective, twenty-two-center, comparative evaluation of renal function in adult surgical patients. Anesth Analg 2000; 90: 683–8Mazze, RI Callan, CM Galvez, ST Delgado-Herrera, L Mayer, DB
Story DA, Poustie S, Liu G, McNicol PL: Changes in plasma creatinine concentration after cardiac anesthesia with isoflurane, propofol, or sevoflurane: A randomized clinical trial. A nesthesiology 2001; 95: 842–8Story, DA Poustie, S Liu, G McNicol, PL
Eger EI II, Gong D, Koblin DD, Bowland T, Ionescu P, Laster MJ, Weiskopf RB: Dose-related biochemical markers of renal injury after sevoflurane versus desflurane anesthesia in volunteers. Anesth Analg 1997; 85: 1154–63Eger, EI Gong, D Koblin, DD Bowland, T Ionescu, P Laster, MJ Weiskopf, RB
Ebert TJ, Messana LD, Uhrich TD, Staacke TS: Absence of renal and hepatic toxicity after four hours of 1.25 minimum alveolar concentration sevoflurane anesthesia in volunteers. Anesth Analg 1998; 86: 662–7Ebert, TJ Messana, LD Uhrich, TD Staacke, TS
Ebert TJ, Frink EJ Jr, Kharasch ED: Absence of biochemical evidence for renal and hepatic dysfunction following 8 hours of 1.25 MAC sevoflurane anesthesia in volunteers. A nesthesiology 1998; 88: 601–10Ebert, TJ Frink, EJ Kharasch, ED
Goldberg ME, Cantillo J, Gratz I, Deal E, Vekeman D, McDougall R, Afshar M, Zafeiridis A, Larijani G: Dose of compound A, not sevoflurane, determines changes in the biochemical markers of renal injury in healthy volunteers. Anesth Analg 1999; 88: 437–45Goldberg, ME Cantillo, J Gratz, I Deal, E Vekeman, D McDougall, R Afshar, M Zafeiridis, A Larijani, G
Sneyd JR: Conflicts of interest: Are they a problem for anaesthesia journals? What should we do about them? Br J Anaesth 2000; 85: 811–4Sneyd, JR
Artusio JF, Van Poznak A, Hunt RE, Tiers FM, Alexander M: A clinical evaluation of methoxyflurane. A nesthesiology 1960; 21: 512–7Artusio, JF Van Poznak, A Hunt, RE Tiers, FM Alexander, M
Summary of the National Halothane Study. JAMA 1966; 197: 775–88NA,
Eger EI II: Dragons and other scientific hazards. A nesthesiology 1979; 50: 1Eger, EI
Higuchi H, Sumita S, Wada H, Ura T, Ikemoto T, Nakai T, Kanno M, Satoh T: Effects of sevoflurane and isoflurane on renal function and on possible markers of nephrotoxicity. A nesthesiology 1998; 89: 307–22Higuchi, H Sumita, S Wada, H Ura, T Ikemoto, T Nakai, T Kanno, M Satoh, T
Higuchi H, Wada H, Usui Y, Goto K, Kanno M, Satoh T: Effects of probenecid on renal function in surgical patients anesthetized with low-flow sevoflurane. A nesthesiology 2001; 94: 21–31Higuchi, H Wada, H Usui, Y Goto, K Kanno, M Satoh, T
Kharasch ED, Frink EJ Jr, Zager R, Bowdle TA, Artru A, Nogami WM: Assessment of low-flow sevoflurane and isoflurane effects on renal function using sensitive markers of tubular toxicity. A nesthesiology 1997; 86: 1238–53Kharasch, ED Frink, EJ Zager, R Bowdle, TA Artru, A Nogami, WM
Kharasch ED, Frink EJ Jr, Artru A, Michalowski P, Rooke GA, Nogami W: Long-duration low-flow sevoflurane and isoflurane effects on postoperative renal and hepatic function. Anesth Analg 2001; 93: 1511–20Kharasch, ED Frink, EJ Artru, A Michalowski, P Rooke, GA Nogami, W
Obata R, Bito H, Ohmura M, Moriwaki G, Ikeuchi Y, Katoh T, Sato S: The effects of prolonged low-flow sevoflurane anesthesia on renal and hepatic function. Anesth Analg 2000; 91: 1262–8Obata, R Bito, H Ohmura, M Moriwaki, G Ikeuchi, Y Katoh, T Sato, S
Ebert TJ, Arain SR: Renal responses to low-flow desflurane, sevoflurane, and propofol in patients. A nesthesiology 2000; 93: 1401–6Ebert, TJ Arain, SR
Eger EI II, Koblin DD, Bowland T, Ionescu P, Laster MJ, Fang Z, Gong D, Sonner J, Weiskopf RB: Nephrotoxicity of sevoflurane versus desflurane anesthesia in volunteers. Anesth Analg 1997; 84: 160–8Eger, EI Koblin, DD Bowland, T Ionescu, P Laster, MJ Fang, Z Gong, D Sonner, J Weiskopf, RB
Fig. 1. Effect of anesthesia on urine protein excretion. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Low-flow (< 2 l/min) sevoflurane data are from surgical patients (n = 120), 15–20 surgical patients with chronic renal insufficiency (n = 56), 1 and normal volunteers (n = 71). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 Comparator data are from low-flow isoflurane in patients with normal renal function (n = 83), 15,17–19 patients with renal insufficiency (n = 53), 1 and volunteers (n = 4) 8,9; high-flow sevoflurane patients (n = 40) 15,16,19; low-flow desflurane patients (n = 18) 20; and propofol patients (n = 10). 20 Box plots show the median, mean (dashed lines  ), 25th and 75th percentiles (box boundaries  ), 10th and 90th percentiles (whiskers  ), and outliers outside the 10th and 90th percentiles. The reference range based on data from healthy nonsurgical subjects is shown by the dotted line. There were no significant differences between the groups (P  = 0.25, Mann–Whitney Rank Sum test).
Fig. 1. Effect of anesthesia on urine protein excretion. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Low-flow (< 2 l/min) sevoflurane data are from surgical patients (n = 120), 15–20surgical patients with chronic renal insufficiency (n = 56), 1and normal volunteers (n = 71). 7–10,21Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10Comparator data are from low-flow isoflurane in patients with normal renal function (n = 83), 15,17–19patients with renal insufficiency (n = 53), 1and volunteers (n = 4) 8,9; high-flow sevoflurane patients (n = 40) 15,16,19; low-flow desflurane patients (n = 18) 20; and propofol patients (n = 10). 20Box plots show the median, mean (dashed lines 
	), 25th and 75th percentiles (box boundaries 
	), 10th and 90th percentiles (whiskers 
	), and outliers outside the 10th and 90th percentiles. The reference range based on data from healthy nonsurgical subjects is shown by the dotted line. There were no significant differences between the groups (P 
	= 0.25, Mann–Whitney Rank Sum test).
Fig. 1. Effect of anesthesia on urine protein excretion. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Low-flow (< 2 l/min) sevoflurane data are from surgical patients (n = 120), 15–20 surgical patients with chronic renal insufficiency (n = 56), 1 and normal volunteers (n = 71). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 Comparator data are from low-flow isoflurane in patients with normal renal function (n = 83), 15,17–19 patients with renal insufficiency (n = 53), 1 and volunteers (n = 4) 8,9; high-flow sevoflurane patients (n = 40) 15,16,19; low-flow desflurane patients (n = 18) 20; and propofol patients (n = 10). 20 Box plots show the median, mean (dashed lines  ), 25th and 75th percentiles (box boundaries  ), 10th and 90th percentiles (whiskers  ), and outliers outside the 10th and 90th percentiles. The reference range based on data from healthy nonsurgical subjects is shown by the dotted line. There were no significant differences between the groups (P  = 0.25, Mann–Whitney Rank Sum test).
×
Fig. 2. Relationship between protein excretion and compound A exposure (area under the curve of inspired compound A concentration vs.  time) during low-flow sevoflurane anesthesia. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Data are from surgical patients (n = 94), 15–20 and normal volunteers (n = 68). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 
Fig. 2. Relationship between protein excretion and compound A exposure (area under the curve of inspired compound A concentration vs. 
	time) during low-flow sevoflurane anesthesia. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Data are from surgical patients (n = 94), 15–20and normal volunteers (n = 68). 7–10,21Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10
Fig. 2. Relationship between protein excretion and compound A exposure (area under the curve of inspired compound A concentration vs.  time) during low-flow sevoflurane anesthesia. Results are from postoperative day 3, which is typically the time of maximum proteinuria. Data are from surgical patients (n = 94), 15–20 and normal volunteers (n = 68). 7–10,21 Data from Eger and from Goldberg are redrawn from figure 3 of Goldberg, multiplying albumin excretion by 1.25 to estimate protein concentration. 10 
×