In modern anaesthesia, low fresh gas flows are common practice in order to reduce costs and pollution as well as to preserve heat and humidify the inspired gas. Increased inspiratory gas concentrations are used to induce and maintain anaesthesia with desflurane and sevoflurane compared with the former volatile anaesthetics. For economic and environmental effects reduced fresh gas flows are especially desirable when these new products are used.
All volatile anaesthetics can be degraded by the lime in the circle absorber system. Desflurane, enflurane and isoflurane react with dry absorbents forming CO . In the case of sevoflurane, several degradation products are generated which are called compounds A–E . The most significant substance is compound A which is a vinylhalide and has nephrotoxic properties with tubular necrosis in rats . In humans there are conflicting results concerning compound A and nephrotoxicity. Patients as well as volunteers were exposed to sevoflurane anaesthesia for up to 8 h, and even longer, with mean compound A concentrations ranging between 20 and 40 p.p.m. [4–9]. There was no clinical evidence of renal injury as judged by changes in blood urea nitrogen (BUN) or creatinine. Elevated sensitive urine markers including proteinuria and glucosuria were found in two investigations but values returned to normal after 1 week [6–8]. Compound A reacts with glutathione forming a S-conjugate which is hydrolysed to the corresponding cysteine S-conjugate. In the kidney, further metabolism occurs by the β-lyase pathway forming a nephrotoxic mercaptan . The cytosolic and mitochondrial β-lyase activity in humans is much less than in rats (1:10 to 1:30) . This could explain the differences in renal involvement in both humans and rats. Until now the critical threshold for compound A toxicity in humans has not been known. Eger [7,8] regards exposure to compound A > 150 p.p.m. h (defined as a compound A concentration multiplied by duration of exposure) to be potentially nephrotoxic in humans.
Compound A production and accumulation in a circle absorber system is dependent on sevoflurane concentration, the type of absorber material (baralyme or soda lime), water content of the absorbent, absorbent temperature, CO2 production and fresh gas flow rates. The construction of the lung ventilator may also influence compound A generation.
Initially, in an attempt to avoid potential toxic effects, the American Food and Drug Administration (FDA) restricted the application of sevoflurane to fresh gas flows > 2 L min−1 but now there is an approval for fresh gas flows of 1 L min−1 for up to 2 h. In the countries of the European Union there is no restriction on the applied fresh gas flows. In minimal flow anaesthesia (0.5 L min−1) higher compound A generation and accumulation can be expected by increasing rebreathing of gas emanating from the absorbent and increasing the amount of CO2 that reaches the absorbent. The aim of this investigation was to evaluate compound A when patients are exposed at fresh gas flow rates of 0.5 L min−1 using inspiratory sevoflurane concentrations up to 3.3 vol% in a Cicero EM (a circle absorber anaesthesia circuit system by Dräger, Lübeck, Germany) and to look for potential side-effects. Due to a special heating system in the Cicero EM increased temperatures are created in the absorber lime . Thus, higher compound A concentrations compared with conventional, unheated absorbers might be expected.
After Institutional Review Board approval and informed consent, American Society of Anesthesiologists physical status I–III patients scheduled for elective surgery (eye, ear, nose and throat and maxillofacial surgery) were randomized to receive minimal flow anaesthesia (0.5 L min−1) either with sevoflurane or isoflurane for at least 2 h. Eligible patients had to be at least 18 years old and without any history of renal or hepatic damage. Patients with elevated liver enzymes (ASAT and ALAT > 200% of the normal range, bilirubin > 3 mg dL−1) and creatinine values (> 1.5 mg dL−1) as well as insulin-dependent diabetes were excluded. General anaesthesia 2 weeks prior to the investigation as well as surgical procedures that might compromise renal blood flow also led to study exclusion.
Anaesthesia was performed in a standardized manner . Fresh soda lime (Drägersorb) was used for every case. Patients were premedicated with 20–40 mg of clorazepate. Anaesthesia was induced with 2–3 mg kg−1 propofol and 0.1–0.2 mg fentanyl after preoxygenation with 5 L min−1 oxygen for 3 min. Rocuronium 0.6 mg kg−1 was administered to facilitate tracheal intubation. Initially, patients were ventilated with a fresh gas flow of 4.4 L min−1 (1.4 L min−1 O2 and 3 L min−1 N2O) using 2.0–2.5 vol% sevoflurane or alternatively 1.0–1.5 vol% isoflurane for maintenance of anaesthesia. After 15 min, fresh gas flow was reduced to 0.5 L min−1 for the duration of the surgical procedure. The inspiratory concentrations of the volatile anaesthetics were adjusted according to the clinical requirements. Supplementary doses of fentanyl (0.1 mg) were given in case of insufficient analgesia. During the whole procedure, inspiratory N2O, O2, inspiratory and expiratory anaesthetic concentrations, tidal volume, minute volume, blood pressure, heart rate, body temperature and oxygen saturation were recorded on-line at 5-min intervals. Compound A concentrations were monitored in patients with sevoflurane anaesthesia. From the inspiratory limb of the breathing circuit, gas samples were taken at defined sampling points using a gas-tight syringe: at baseline before start of sevoflurane, immediately before flow reduction, every 30 min after flow reduction up to 2 h, afterwards every hour and immediately before wash-out period.
Venous blood and urine samples were taken the day before anaesthesia and on days 1, 3 and 5 after anaesthesia. Blood samples were analysed for creatinine, urea, total protein, glucose, bilirubin and ASAT and ALAT. Urine sediment was microscopically analysed. With rapid semi-quantitative urine tests pH, osmolality, protein and glucose were determined. In a subgroup of 32 patients (random number 17–48) quantitative urine measurements were made. The collection period was 12 h prior to anaesthesia and on the day of surgery from 20:00 to 08:00 h and for 24 h on days 2 and 4 from 08:00 to 20:00 h. Measurements of creatinine, glucose, protein, albumin, IgG, α1-microglobulin and α- and π-GST (Glutathion-S-Transferase) were performed to detect the presence of renal cytotoxic processes.
All chemical analyses were made by the central laboratories of the University of Münster. Compound A concentrations were determined by gas chromatography with flame ionization detection at the University of Mainz using the method previously described by Cunningham .
Anaesthetic and compound A exposure were calculated from the area under the curve of the inspiratory concentration vs. time (vol% h, p.p.m. h). Values are presented as mean ± standard deviation. Patients' demographic data were analysed using the Student's unpaired t-test. Inter- and intragroup comparisons of laboratory data were performed using two-way repeated measures analysis of variance. Statistical significance was assigned at P < 0.05.
Ninety patients were randomized, 87 patients underwent study protocol. Eleven patients had to be excluded because of protocol violations (in three patients the fresh gas flow 0.5 L min−1 < 2 h, one patient had elevated liver enzymes preoperatively and one patient had incomplete follow up) or faulty compound A measurements (six patients). Thus, a total of 76 patients (33 sevoflurane and 43 isoflurane) were available for complete evaluation. Table 1 shows patients' demographic data while Table 2 represents data during anaesthesia. No difference was found with respect to concomitant treatment (antibiotics, etc.).
Figure 1 shows the mean compound A concentration and the corresponding mean sevoflurane concentrations. After 90 min of minimal flow anaesthesia, a plateau phase was reached and no further rise in concentration of compound A was found. Mean maximal compound A concentration was 40 ± 9 p.p.m. with a mean corresponding inspiratory sevoflurane concentration of 2.1 ± 0.5 vol%. During the plateau phase there was a strong correlation between the concentrations of compound A and sevoflurane (Figure 2).
Table 3 shows the blood chemistry during the whole study period. There is no difference between groups and no difference within groups compared with baseline. The mean serum values of creatinine, urea, protein, glucose, bilirubin, ASAT and ALAT do not reveal any hepatic or renal derangement by the different volatile anaesthetics.
Mean values of quantitative urine measurements are shown in Table 4. High standard deviations indicate that the wide spread distribution of values. Data for 15 sevoflurane and 16 isoflurane patients were available. As there were two different collection periods, 12 h and 24 h, and to exclude errors in urine sample collection (spontaneous urine sampling) values were referred to urine creatinine excretion. No differences in protein excretion were found between patients with sevoflurane and isoflurane anaesthesia. There was an increase in protein, α1-microglobuline and α-GST excretion for both groups compared with baseline. Scatterplots (Figures 3–5) give a more precise idea of the distribution of the values obtained at the different points of measurement. Semi-quantitative urine measurements and microscopic findings of the sediment did not reveal any renal involvement.
This study was designed to investigate the effects of minimal flow anaesthesia (0.5 L min−1 fresh gas flow) with sevoflurane in a Cicero EM anaesthesia system during clinical conditions. Only small amounts of fentanyl were used and anaesthesia was maintained with the volatile anaesthetic. Until now no data concerning minimal flow (0.5 L min−1) anaesthesia with sevoflurane has been available. The main question is ‘what compound A concentrations are generated under these conditions, and does compound A accumulate under prolonged minimal flow?’ Second, the side-effects of compound A exposure were studied by blood and urine biochemistry.
A fresh gas flow of 0.5 L min−1 led to an increased inspiratory concentration of compound A. Clinical studies with fresh gas flows at 1 L min−1 [4,5] have shown mean maximum compound A concentrations < 30 p.p.m. compared with 40 p.p.m. in this investigation. Correlation between sevoflurane and compound A concentrations showed a steeper slope compared with data obtained at 1 L min−1 fresh gas flow in previously published studies [4,6]. This effect might be due to the reduction in fresh gas flow but also to the special heating system in the Cicero EM system which creates high temperatures in the absorber . In general, the results of this study cannot be transferred to different settings at fresh gas flows of 0.5 L min−1 with different anaesthetic ventilator systems. Data obtained with a Physioflex® system (a closed circuit anaesthesia machine) resulted in lower compound A concentration compared with our data obtained in the present study [15–17]. The circuit gases in the Physioflex® are circulated unidirectionally by a blower at a rate of 70 L min−1, which effectively reduces the temperature in the absorber thus resulting in a reduction in compound A concentration. We found that after 90 min of flow reduction there was no further rise in concentration of compound A and a plateau phase was reached. The inspiratory concentration of compound A strongly correlated with the sevoflurane concentrations. Not only are the absolute values of the inspiratory compound A concentrations but the time of compound A exposure seems to be important when comparing different studies and potential toxic side-effects reported. As a result of a shorter mean time of anaesthesia 3 h vs. 6.7 h the mean compound A exposure (p.p.m. h−1) was nearly 50% less in this study compared with the results of Higuchi and his colleagues . The range of values obtained for compound A concentrations in the Higuchi study (8.6–56.9 p.p.m.) was similar to values in this study (19–57 p.p.m.) . Compound A exposure (102 ± 33 p.p.m. h−1) in the current study was similar to the studies published by Kharash and his colleagues  and Bito and his colleagues . In both studies renal involvement was not found [4,5]. From a study of volunteers, Eger suggested a critical threshold for compound A exposure to be about 150 p.p.m. h−1 [7,8]. After 8-h sevoflurane anaesthesia (3 Vol.%) with a fresh gas flow of 2 L min−1, transient albuminuria, glucosuria, α-and π-GST elevations were found by Eger's group . Ebert's study could not reproduce these findings in a similar setting . Differences in these studies might be due to different anaesthesia ventilator systems. Most authors do not mention the anaesthetic apparatus used in their studies. Higuchi and his colleagues found a mean exposure of compound A of 192 ± 46 p.p.m. h−1 transient proteinuria which correlated with concentrations of compound A exposure . Concentrations of N-acetyl-glucosaminidase (NAG) were found raised during both high- and low-flow sevoflurane anaesthesia and these values correlated with serum fluoride concentrations . In all these studies renal involvement was not evident by changes in serum creatinine and urea values.
Different urine sampling periods were used in this study (12 and 24 h). Excreted urine creatinine was used as a urine marker to compare results from different sampling periods and to eliminate errors from improper spontaneous urine collection . There were no differences in renal effects comparing sevoflurane and isoflurane anaesthesia. The measured urine markers and the microscopic findings of the urine sediment did not reveal any tubular damage by sevoflurane and compound A in this study. Increases in protein, α1-microglobuline and α-GST excretion compared with baseline were found for sevoflurane and isoflurane anaesthesia in the postoperative course. Stress and factors related to surgery may have induced these changes as well as intraindividual variation . Until now it has not been easy to assess renal damage by specific markers . Even for established markers such as albumin and NAG, our knowledge is incomplete and this limits the diagnostic value of urine markers for monitoring renal damage . However, if renal markers are not elevated renal damage can be excluded. The predictive value of elevated markers is unknown and does not correlate to specific injuries and their course . In healthy people, orthostasis, hypertension or cigarette smoking may induce proteinuria without any pathological significance . These facts might explain the different findings in different studies. Unless there are better parameters, renal involvement can only be assessed by measuring serum creatinine and urea. However, these parameters are not very sensitive in detecting minor changes in renal function. Investigations in patients with impaired renal function may provide more information about renal involvement by sevoflurane and compound A. In the former absorbents, sodium hydroxide or potassium hydroxide appear to enhance the production of compound A and carbon monoxide by degrading volatile anaesthetics . New absorbents containing calcium hydroxide or lithium hydroxide could eliminate any potential hazards from the toxic compounds by decreasing the production of compound A and carbon monoxide [20,21].
In conclusion, we did not find deleterious effects of minimal flow sevoflurane anaesthesia (0.5 L min−1 fresh gas flow) for at least 2 h. Patients were clinically surveyed and biochemical tests were performed to detect potentially harmful effects of sevoflurane and compound A respectively. Minimal flow anaesthesia led to enhanced accumulation of compound A. As compound A is not measured in clinical practice and a general toxicological profile in humans has been uncertain to date, further investigations are needed to evaluate compound A generation and accumulation in different anaesthesia breathing and ventilator systems under varying conditions and to assess the individual margin of safety.
Source of financial support: Abbott GmbH, Max-Planck-Ring 2, 65205 Wiesbaden, Germany.
1 Fang ZX, Eger II EI, Laster MJ, Chortkoff BS, Kandel L, Ionescu P. Carbon monoxide production from degradation of desflurane, enflurane, isoflurane, halothane and sevoflurane by soda lime and baralyme. Anesth Analg
1995; 80: 1187–1193.
2 Cunningham DD, Huang S, Webster J, Mayoral J, Grabenkort RW. Sevoflurane degradation to compound A
breathing systems. Br J Anaesth
1996; 77: 537–543.
3 Keller KA, Callan C, Prokocimer P et al. Inhalation
toxicology study of haloalkene degradant of sevoflurane, Compound A
(PIFE) in Sprague–Dawley rats. Anesthesiology
1995; 83: 1220–1232.
4 Kharash ED, Frink Jr EJ, 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. Anesthesiology
1997; 86: 1238–1253.
5 Bito H, Ikeuchi Y, Ikeda K. Effects of low-flow sevoflurane on renal function. Comparison with high-flow sevoflurane and low-flow isoflurane anesthesia. Anesthesiology
1997; 86: 1231–1237.
6 Higuchi H, Sumita S, Wada H et al.
Effects of sevoflurane and isoflurane on renal function and on possible markers of nephrotoxicity. Anesthesiology
1998; 89: 307–322.
7 Eger II EI, Koblin DD, Bowland T et al.
Nephrotoxicity of sevoflurane versus desflurane anesthesia in volunteers Anesth Analg
1997; 84: 160–168.
8 Eger II EI, Gong D, Koblin DD et al.
Dose -related biochemical markers of renal injury after sevoflurane versus desflurane anesthesia in volunteers Anesth Analg
1997; 85: 1154–1163.
9 Ebert TJ, Frink EJ, Jr Kharasch ED. Absence of biochemical evidence for renal and hepatic dysfunction after 8 hours of 1.25 minimum alveolar concentration sevoflurane anesthesia in volunteers. Anesthesiology
1998; 88: 601–610.
10 Iyer RA, Frink EJ, Ebert TJ, Anders MW. Cysteine conjugate β-lyase-dependent metabolism of compound A
(2-[fluoromethoxy]-1,1,1,3,3,3-pentafluoro-1-propene) in human subjects anesthetized with sevoflurane and in rats given compound A
1998; 88: 611–618.
11 Iyer RA, Anders MW. Cysteine conjugate β-lyase-dependent biotransformation of cysteine S-conjugates of the sevoflurane degradation product compound A
in human, nonhuman primate, and rat kidney cytosol and mitochondria. Anesthesiology
1996; 85: 1454–1461.
12 Kuhn I, Wissing H. Atemgasklimatisierung und Klimaprofil des Narkosegerätes Dräger Cicero. Anästhesiol Intensivmed Notfallmed Schmerzther
1998; 33 (Suppl. 3): S225.
13 Baum JA. Low Flow Anaesthesia
Oxford: Butterworth, Heinemann, 1996: 169–193.
14 Cunningham DD, Webster J, Nelson D, Williamson B. Analysis of sevoflurane degradation products in vapor phase samples. J Chromatogr B Biomed Appl
1995; 668: 41–52.
15 Rolly G, Versichelen L, Bouche MP et al.
In vitro analysis of compound A
formation by sevoflurane in a computer controlled closed circuit Physioflex anesthesia apparatus Anesth Analg
1999; 88: S193.
16 Funk W, Gruber M, Jakob W, Hobbhahn J. Compound A
does not accumulate during closed circuit sevoflurane anaesthesia
with the Physioflex. Br J Anaesth
1999; 83: 571–575.
17 Bito H, Suzuki A, Sanjo Y, Katoh T, Sato S. Comparison of compound A
with sevoflurane anaesthesia
using a closed system with a Physioflex anaesthesia
machine vs. a low-flow system with a conventional anaesthesia
machine. Br J Anaesth
2000; 84: 350–353.
18 Hofmann W, Edel H, Guder WG. Urineiweiβ-Differenzierung. Ein neues Konzept zur nichtinvasiven Diagnostik von Nierenerkrankungen. Münch Med Wochenschr
1997; 139: 488–494.
19 Baines AD. Strategies and criteria for developing new urinalysis tests. Kidney Intern
1995; 46 (Suppl. 47): S137–S141.
20 Bedford RF, Ives HE. The renal safety of sevoflurane. Anesth Analg
2000; 90: 505–508.
21 Stabernack C, Brown R, Laster MJ, Dudziak R, Eger II EI. Absorbents differ enormously in their capacity to produce compound A
and carbon monoxide. Anesth Analg