Sevoflurane is degraded during low-flow anesthesia to fluoromethyl-2,2-difluoro-1-(trifluoromethyl)vinyl ether ("compound A"), which causes renal necrosis in rats but is not known to cause nephrotoxicity in surgical patients. Compound A is metabolized to glutathione S-conjugates and then to cysteine S-conjugates, which are N-acetylated to mercapturic acids (detoxication pathway), or metabolized by renal beta-lyase to reactive intermediates (toxification pathway) and excreted as 3,3,3-trifluoro-2-fluoromethoxypropanoic acid. This investigation quantified compound A metabolites in urine after low-flow sevoflurane administration, to assess relative flux via these two pathways.
Patients (n = 21) with normal renal function underwent low-flow (11 min) sevoflurane anesthesia designed to maximize compound A formation. Inspiratory, expiratory, and alveolar compound A concentrations were quantified. Urine mercapturic acids and 3,3,3-trifluoro-2-fluoromethoxypropanoic acid concentrations were measured by gas chromatography and mass spectrometry.
Sevoflurane exposure was 3.7 +/- 2.0 MAC-h. Inspired compound A maximum was 29 +/- 14 ppm; area under the inspired concentration versus time curve (AUCinsp) was 78 +/- 58 ppm x h. Compound A dose, calculated from pulmonary uptake, was 0.39 +/- 0.35 mmol (4.8 +/- 4.0 micromol/kg) and correlated with AUCinsp (r2 = 0.84, P < 0.001). Mercapturic acids excretion was complete after 2 days, whereas 3,3,3-trifluoro-2-fluoromethoxypropanoic acid excretion continued for 3 days in some patients. Total (3-day) mercapturates and fluoromethoxypropanoic acid excretion was 95 +/- 49 and 294 +/- 416 micromol, respectively (1.2 +/- 0.6 and 3.6 +/- 5.0 micromol/kg).
Compound A doses during 3.7 MAC-h, low-flow sevoflurane administration in humans are substantially less than the threshold for renal toxicity in rats (200 micromol/kg). Compound A metabolites quantification may provide a biomarker for compound A exposure and relative metabolism via toxification and detoxication pathways. Compared with previous investigations, relative metabolic flux (fluoromethoxypropanoic acid/mercapturates) through the toxification pathway was sixfold greater in rats than in humans. Species differences in dose and metabolism may influence compound A renal effects.
SEVOFLURANE undergoes base-catalyzed degradation in carbon dioxide absorbents to the haloalkene fluoromethyl-2,2-difluoro-1-(trifluoromethyl)vinyl ether (“compound A”), which occurs in higher amounts during low-flow anesthesia. 1Typical of numerous haloalkenes, 2,3compound A causes nephrotoxicity in rats, specifically corticomedullary proximal tubular necrosis, accompanied by proteinuria, glucosuria, and enzymuria. 4,5The threshold for renal tubular necrosis in rats (250 g) is generally accepted to be 290–340 ppm · h. 5–7There are, however, species differences in the threshold for compound A nephrotoxicity. Monkeys showed histologic and biochemical evidence of renal toxicity at 800 ppm · h but not 600 ppm · h. 8Swine showed no renal tubular cell necrosis after 612 ppm · h, and a threshold for toxicity has not been established. 9In humans, numerous investigations have evaluated renal function after surgical anesthesia with low-flow sevoflurane, in which maximum inspired compound A concentrations averaged 8–24 and 20–32 ppm with soda lime and barium hydroxide lime, respectively, and exposures were as high as 400 ppm · h. 10–17Using standard clinical (creatinine clearance, serum creatinine, and BUN) and experimental (proteinuria, glucosuria, and enzymuria) markers of renal functional and structural integrity, investigations to date have found no evidence of renal toxicity after administration of low-flow anesthesia, 10–15although one 16but not another 17found mild, transient proteinuria, suggesting altered glomerular permeability.
Like numerous other haloalkenes, 2,3compound A undergoes glutathione conjugation, with renal uptake and metabolism of S -conjugates (fig. 1). There are extensive qualitative similarities in compound A metabolism between rats and humans. In rats, compound A undergoes nonenzymatic and glutathione S -transferase catalyzed metabolism to two alkane (diastereomers of S -[1,1,-difluoro-2-fluoromethoxy-2-(trifluoromethyl)ethyl]-glutathione) and two alkene ((E )- and (Z )-S -[1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl]glutathione) glutathione conjugates, which undergo cleavage to the corresponding cysteine S- conjugates. 18–20Cysteine S- conjugates N -acetylation forms the mercapturic acids N -acetyl-S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine and (E )- and (Z )-N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine, which are excreted in rat urine. 18,20,21These mercapturic acids have also been qualitatively identified in human urine after administration of low-flow sevoflurane, 22indicating that compound A glutathione- and cysteine S -conjugate formation also occurs in humans. Compound A cysteine S- conjugates are metabolized by rat renal cysteine conjugate β-lyase in vitro and in vivo to reactive intermediates that may bind to cellular macromolecules or undergo hydrolysis to 3,3,3-trifluoro-2-fluoromethoxypropanoicacid. 20,22–24Iyer et al. reported that 3,3,3-trifluoro-2-fluoromethoxypropanoic acid could decompose to trifluorolactic acid. 24Identification of 3,3,3-trifluoro-2-fluoromethoxypropanoic acid and trifluorolactic acid in urine after low-flow sevoflurane administration 22indicated that compound A cysteine S -conjugates can also undergo β-lyase-catalyzed metabolism in humans.
Compound A and other haloalkene cysteine S -conjugate metabolism by renal β-lyase is considered a toxification pathway, whereas N -acetylation and mercapturic acid formation is regarded as a detoxication pathway. 7,18–25Although compound A mercapturates and β-lyase–dependent metabolites have been qualitatively identified in human urine, they have not been quantified, and the relative metabolic flux through these toxification and detoxication pathways in humans has not been determined. In addition, interspecies differences in the relative metabolic flux through these pathways may confer differences in susceptibility to compound A nephrotoxicity. Therefore the purpose of this investigation was to provide GC/MS identification of compound A mercapturates, 3,3,3-trifluoro-2-fluoromethoxypropanoic acid, and trifluorolactic acid in human urine, and to quantify metabolite excretion reflecting relative compound A toxification and detoxication in humans undergoing low-flow sevoflurane anesthesia.
Materials and Methods
Twenty-one patients who were American Society of Anesthesiologists physical status I–III, without history of hepatic or renal disease, and undergoing anesthesia for elective noncardiac and nonaortic surgery with planned duration > 2 h were studied. Patients (5 men, 16 women) aged 49 ± 12 yr (range, 30–69 yr) and weighing 76 ± 12 kg (range, 52–97) were studied. The investigation was approved by the Institutional Review Board, and all patients provided written informed consent. The anesthetic protocol was designed to maximize compound A formation (fresh barium hydroxide lime, total gas flow rate 1 l/min, and high sevoflurane concentrations, achieved by precluding opioids [except 50–150 μg fentanyl for induction], nitrous oxide, and intraoperative neuraxial opioids and local anesthetics). End-tidal anesthetic concentrations were monitored continuously (Capnomac, Datex Medical Instrumentation, Tewksbury, MA). Respiratory gas samples were obtained using gas-tight syringes, from the inspiratory (FI) and expiratory (FE) limbs of the anesthesia circuit adjacent to the respective valves, and from a catheter positioned coaxially in the endotracheal tube with the tip protruding just beyond the tube orifice after ventilation was held for 30 s (alveolar samples, FA). Samples were obtained after intubation, 5, 10, 15, 30, 60, 90, and 120 min after the start of low-flow anesthesia, hourly thereafter, and at the end of low-flow anesthesia. Urine was collected in 24-h intervals for 72 h after anesthesia. The volume was measured, and an aliquot was frozen at −80°C for later analysis. Additional details regarding the experimental protocol, and results of renal function evaluations in a multicenter superset of these patients, have been published previously. 15Data in this investigation represent additional evaluations performed at the time of the original investigation, and subsequent laboratory studies, on the University of Washington cohort from the previous investigation.
Compound A concentrations in respiratory gas samples were determined by gas chromatography with flame ionization detection, as described previously. 15
Urine concentrations of N -acetyl-S -(1,1,3,3,3-pentafluoro-2-(fluoromethoxy)propyl)-L-cysteine, N -acetyl-S -(1-fluoro-2-(fluoromethoxy)-2-(trifluoromethyl)-vinyl)-L-cysteine, 3,3,3-trifluoro-2-(fluoromethoxy)propanoic acid, and trifluoro-lactic acid were determined by GC/MS as described previously. 26Briefly, the internal standards N -acetyl-S -(2,2-di-fluro-vinyl)-L-cysteine (1.25 μg) or dichloroacetic acid (150 ng) were added to urine (0.1–1 ml) that was acidified and extracted with diethyl ether. Samples were derivatized with diazomethane or diphenyldiazomethane for mercapturic acid or fluoropropionic acid analysis, respectively, and analyzed by GC/MS. Analyses were performed on a Hewlett-Packard 5890 Series II GC-5972 mass selective detector, using a DB-17 column (30 m x 0.32 mm x 0.5 μ)(J & W Scientific, Folsom, CA). The injector and detector temperatures were 250°C and 300°C, respectively, and the column head pressure was 5 psi for all assays. For mercapturates analysis, the oven was at 35°C for 5 min, increased 5°C/min to 210°C, then at 20°C/min to 280°C and held for 7 min. Selected-ion monitoring was used to quantify the methyl esters of N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine (m/z 236), N -acetyl-S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine (m/z 256), and N -acetyl-S -(2,2-difluorovinyl)-L-cysteine (m/z 196). For analysis of 3,3,3-trifluoro-2-fluoromethoxypropanoic acid and trifluorolactic acid, the oven was at 35°C for 5 min, increased 20°C/min to 250°C, then at 10°C/min to 280°C, and held for 10 min. Selected-ion monitoring was used for diphenylmethyl esters of 3,3,3-trifluoro-2-fluoromethoxypropanoic (m/z 342), trifluorolactic (m/z 310), and dichloroacetic(m/z 294) acids.
Calibration standards containing N -acetyl-S -(1,1,3,3,3-pentafluoro-2-(fluoromethoxy)propyl)-L-cysteine (0.5–116 μg/ml), N -acetyl-S -(1-fluoro-2-(fluoromethoxy)-2-(trifluoromethyl)-vinyl)-L-cysteine (0.5–134 μg/ml), 3,3,3-trifluoro-2-(fluoromethoxy)propanoic acid (1–50 μg/ml), and trifluorolactic acid (0.1–10 μg/ml) were prepared daily using blank urine. Standard curves of peak area ratios were linear over the concentrations used (r2> 0.99 for all analytes), and the limit of quantification was defined as the lowest point on the standard curve. Interday coefficients of variation were:N -acetyl-S -(1,1,3,3,3-pentafluoro-2-(fluoromethoxy)propyl)-L-cysteine (6% and 5% at 0.9 and 46 μg/ml), N -acetyl-S -(1-fluoro-2-(fluoromethoxy)-2-(trifluoromethyl)-vinyl)-L-cysteine (6% and 2% at 1.1 and 54 μg/ml), 3,3,3-trifluoro-2-(fluoromethoxy)propanoic acid (6% at 2.5 and 25 μg/ml), and trifluorolactic acid (6% and 4% at 0.5 and 25 μg/ml).
Sevoflurane exposure was calculated as the area under the end-tidal concentration versus time curve, determined in 15-min intervals. Compound A exposure was similarly calculated as the product of inspiratory concentration and time. To permit comparison with the previously published data, 15the 5-, 10-, and 15-min samples were not used for this calculation. The dose of compound A actually administered was taken as compound A uptake, calculated as described previously for sevoflurane and other anesthetics. 27Uptake rate was calculated as
where VUwas the total pulmonary uptake rate (L compound A vapor/min), VEwas the minute ventilation (l/min), and FIand FMwere the inspired and mixed expired compound A concentrations, respectively, determined at each measurement interval. FMwas calculated according to
where fAand fDrepresent the fraction of alveolar and dead space ventilation, respectively. Values for fAand fDduring mechanical ventilation were taken as 0.5 each. 28Mixed-expired gas samples are conventionally obtained by placing a mixing box in the expiratory limb of the anesthesia circuit. Preliminary studies showed that sevoflurane concentrations in mixed expired samples obtained distal to a 3-l mixing box were not substantially different than those measured at the end of the circuit adjacent to the expiratory valve without the box (not shown); thus, the mixing box was not subsequently used to obtain mixed-expired compound A samples. VUwas therefore also calculated using FEfor FM, which provided an upper bound (more conservative estimate) of VU. Values for total pulmonary compound A uptake (l/min) were converted to mol/min by application of the general gas equation. 27The sum of the products of pulmonary compound A uptake rate and exposure time for each interval gave the total dose in moles.
Results are expressed as the mean ± SD. Correlations were analyzed by linear regression analysis. Statistical significance was assigned at P < 0.05.
Compound A Concentrations
Compound A was readily detected at the first sampling time, 5 min after the start of low-flow anesthesia. Mean concentrations are shown in figure 2A. Concentrations in most patients plateaued after approximately 30 min (shown previously 15). They continued to increase in three patients, two of whom underwent laparoscopic procedures with CO2insufflation 29and who also received much higher sevoflurane concentrations (3.0%, 3.1%, and 2.6%) compared with all others (mean, 1.9%), which influenced the mean values. Expired and alveolar compound A concentrations were significantly correlated (slope = 0.77, r2= 0.75, P < 0.001, data not shown), although there was greater variability in the latter measurement (fig. 2A), possibly reflecting the fact that mechanical ventilation, but not fresh gas flow, was interrupted during alveolar sampling. Compound A uptake was approximated from the FE:FIratio (fig. 2B). The ratio rapidly reached equilibration; 5-min values were 91% and 87% of those at 30 and 60 min, respectively, although the equilibrium ratio was only 0.64 at 2 and 3 h. Similar results were obtained using the more conventional FA:FIratio, although variability was greater because of the greater variability in FA(not shown).
Compound A exposure and estimated dose are provided in table 1. Exposure, measured as the inspired AUC (AUCinsp), was 78 ± 58 ppm · h (range, 10–223). Total dose, estimated from the pulmonary uptake rate, and conventionally calculated using FAand FIto determine FM, was 0.23 ± 0.22 mmol. Uptake was also calculated using FEas the value for FM(0.39 ± 0.35 mmol; 4.8 ± 4.0 μmol/kg). This method avoids potential sampling confounders, makes no assumptions about the proportions of alveolar and dead space ventilation, 28does not consider any compound A adsorption by the anesthesia circuit, 30and provides a more conservative (higher) estimate for dose. There was a significant correlation (r2= 0.84, P < 0.001) between this dose and compound A exposure measured as AUCinsp(fig. 3A) or as AUCinsp-exp(r2= 0.78, P < 0.001, data not shown). A similar correlation was also observed (r2= 0.84) using the lower estimate for compound A dose and AUCinsp(not shown). The higher estimate for compound A dose (0.39 ± 0.35 mmol) is used in the remainder of this report. There was also a significant correlation (r2= 0.76, P < 0.001) between compound A dose and sevoflurane exposure (MAC-h;fig. 3B).
Compound A Metabolites Excretion
Excretion of the mercapturic acids N -acetyl-S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine and N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine in the urine of patients exposed to compound A during low-flow sevoflurane was identified and quantified. An extracted, methylated urine sample (fig. 4A) showed peaks at 35.2 and 35.5 min that were not present in preanesthesia urine (not shown). Mass spectra of these peaks are shown in figures 4B and 4C. Diagnostic ions and corresponding fragments were m/z 278 ([M-COOCH3]+) and 236 ([M-COOCH3-COCH3]+) for the alkene;m/z 298 ([M-COOCH3]+), 256 ([M-COOCH3-COCH3]+) and 176 ([C6H10NO3S]+) for the alkane. Characteristic mercapturate fragments (m/z 144, C5H6NO2S and 88 C3H6NO2) were also observed. Based on similarities of retention time and mass spectra to those of the synthetic compound A mercapturic acids, 26the excreted compounds were identified as N -acetyl-S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine and N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine. No attempts were made to resolve and separately quantify the (E )- and (Z )-isomers of N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)-vinyl)-L-cysteine.
Excretion of 3,3,3-trifluoro-2-fluoromethoxypropanoic acid, which results from renal β-lyase–mediated metabolism of compound A cysteine S -conjugates, was also identified and quantified. An extracted, derivatized urine sample (fig. 5A) showed a peak at 16.4 min that was not present in urine obtained before anesthesia (not shown). The mass spectrum of this peak (fig. 5B) is identical to that of the synthetic, derivatized acid (not shown). 26Concentrations ranged from 0.5 to 79 μg/ml. 3,3,3-Tri-fluoro-2-fluoromethoxypropanoic acid has been reported to decompose to trifluorolactic acid in vitro and in vivo , 21,24and excretion of this acid would also represent compound A cysteine S -conjugates metabolism by β-lyase. Trifluorolactic acid was detected in urine (fig. 5A, 16.2 min), based on selected-ion monitoring, although concentrations were low (0–1.1 μg/ml), highly variable, and insufficient to obtain full scan mass spectra of the excreted compound. Identification, therefore, was based on selected-ion monitoring of diagnostic ions, as described previously. 21
Daily excretion of mercapturic acids and β-lyase–derived fluoroacid metabolites in urine after low-flow sevoflurane administration is described in table 2. Mercapturic acids were readily detected for 2 days postoperatively in all subjects, and 3,3,3-trifluoro-2-fluoromethoxypropanoic acid was detected in all patients on day 1, and in approximately half of the patients by day 3. In contrast, trifluorolactic acid concentrations were above the limit of quantification in only five subjects on day 1 and in only two subjects on day 2. Cumulative 3-day metabolite excretion is shown in figure 6. Mercapturic acids excretion was complete after 2 days, whereas additional 3,3,3-trifluoro-2-fluoromethoxypropanoic acid continued for 3 days in some patients. There were significant linear correlations between compound A dose and total 72-h excretion of mercapturic acids (r2= 0.64, P < 0.001), β-lyase–derived fluoroacid metabolites (r2= 0 .55, P < 0.001), and total metabolite excretion (r2= 0.59, P < 0.001; data not shown).
Compound A Exposure
This investigation was designed, in part, to provide a more thorough characterization of compound A disposition than conventionally obtained in clinical studies, in which only inspired concentrations are measured, and usually beginning 0.5–2 h after initiation of low-flow anesthesia. 10–16,31–35Some of these more recent investigations have calculated compound A exposure (AUCinsp) using the trapezoidal rule from the origin to the first measured time. We detected compound A 5 min after the start of low-flow anesthesia, and at concentrations greater than would have been predicted by extrapolating linearly from time zero to the first 0.5–2 h measurement. One other investigation also detected compound A immediately after the onset of low-flow anesthesia. 36These results show that many previous investigations have somewhat underestimated actual compound A exposure, with the underestimate greater for those measuring compound A beginning at 2 h. 34,35
Compound A uptake, estimated from the FE/FIratio, was rapid. The FE/FIratio of 0.64 observed presently is similar to that previously described by Frink et al. 10,32,33It is also consistent with the FA/FIratio of 0.8 reported to be achieved after 1–2 h, in which end-tidal gas samples were used to estimate alveolar compound A concentrations. 36The rapid increase in FE(A)/FIratio is consistent with the low solubility of compound A (blood:gas partition coefficient of 0.31);37however, the equilibrium value is lower than the solubility would predict. As identified earlier, 36this may relate to compound A reactivity with blood and plasma constituents, 37most likely protein thiols, 18,19and possibly also reactivity with tissue thiols. 19,38
The dose of compound A taken up can be compared with exposure to compound A and to sevoflurane. Compound A exposure (measured as AUCinsp) was an excellent predictor of compound A dose (measured from inspired and expired compound A concentrations and minute ventilation). Low-flow sevoflurane exposure (end-tidal MAC-h) was also an excellent predictor of the compound A dose (approximately 0.1 mmol compound A per MAC-h low-flow [1 l/min] sevoflurane). These relationships can be used to estimate compound A doses in other investigations in which it is not measured directly.
Low-flow sevoflurane in this investigation (3.7 ± 2.0 MAC-h) resulted in a compound A dose of 0.39 ± 0.35 mmol. Previously, a similar sevoflurane exposure (3.7 ± 0.3 MAC-h) resulted in a sevoflurane dose of 88.8 ± 28.8 mmol. 27Assuming that the same sevoflurane exposure in the present investigation resulted in similar sevoflurane uptake, the compound A dose was approximately 1/200 that of sevoflurane. This finding is consistent with several investigations showing that inspired compound A concentrations are approximately 1/500–1/1,000 those of sevoflurane, 12,31and with the solubility and FE(A)/FIratio for compound A.
Compound A doses administered during low-flow anesthesia in humans can be compared with those causing nephrotoxicity in rats. Although animal investigations using inhalation exposure do not permit assessment of the actual compound A dose administered, 4–7doses are known when intraperitoneal injection is used. 25The nephrotoxic threshold in rats was 200 μmol/kg intraperitoneal compound A, 25which produced histologic and biochemical alterations similar to those elicited by the inhaled threshold of 290–340 ppm · h. 5–7The compound A dose received by patients undergoing 3.7 ± 2.0 MAC-h low-flow sevoflurane (4.8 ± 4.0 μmol/kg) is substantially less than the nephrotoxic threshold in rats (200 μmol/kg).
Compound A Metabolism
Results of this investigation provide clear GC/MS identification and quantitation of compound A alkane and alkene mercapturic acids and 3,3,3-trifluoro-2-fluoromethoxypropanoic acid excretion in urine. They confirm qualitative NMR and selected-ion mode GC/MS identifications published previously. 22These results demonstrate that compound A undergoes metabolism in humans via a complex bioactivation scheme involving the concerted participation of several organs and enzyme systems, collectively referred to as the β-lyase pathway. 2,3Specifically, identification of the mercapturic acids N -acetyl-S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine and N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine demonstrates that compound A undergoes metabolism to the glutathione S -conjugates S -[1,1,-difluoro-2-fluoromethoxy-2-(trifluoromethyl)ethyl]-glutathione) and S -[1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl]glutathione, which in turn are cleaved to the corresponding cysteine S- conjugates S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine and S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine, which undergo subsequent N -acetylation. Slightly greater amounts of N -acetyl-S -(1,1,3,3,3-pentafluoro-2-fluoromethoxypropyl)-L-cysteine compared with N -acetyl-S -(1-fluoro-2-fluoromethoxy-2-(trifluoromethyl)vinyl)-L-cysteine were excreted. It is not known, at present, whether this represents differences in the biosynthesis of the alkane and alkene glutathione S -conjugates, or in their subsequent interorgan transport and processing. Identification of 3,3,3-trifluoro-2-fluoromethoxypropanoic acid demonstrates that the cysteine S- conjugates undergo metabolism by renal cysteine conjugate β-lyase in vivo to reactive intermediates, which are subsequently hydrolyzed. Because 3,3,3-trifluoro-2-fluoromethoxypropanoic acid is a common product of both S- conjugates metabolism, 20,22–24the present results do not indicate whether one, or both, cysteine S- conjugates are metabolized by β-lyase. 3,3,3-Trifluoro-2-fluoromethoxypropanoic acid has been reported to be unstable, degrading to trifluorolactic acid. 24Quantification of this degradation product in vivo might be important, because it would also represent cysteine S- conjugates metabolism by β-lyase. Trifluorolactic acid was rarely observed, however, and never accounted for more than a small fraction of the 3,3,3-trifluoro-2-fluoromethoxypropanoic acid excreted in human urine. Assay sensitivity does not appear to explain this result, as recovery averaged 95% and 0.1μg/ml was readily detectable. Trifluorolactic acid excretion in human urine after low-flow sevoflurane was identified previously. 22Excretion was not quantified, but qualitatively appeared small relative to that of the mercapturates and 3,3,3-trifluoro-2-fluoromethoxypropanoic acid. 22Thus, although 3,3,3-trifluoro-2-fluoromethoxypropanoic acid decomposition to trifluorolactic acid readily occurred in vitro , 24this finding does not appear quantitatively significant in humans in vivo .
Relative cysteine S -conjugates metabolism by β-lyase versus N -acetylation can be assessed by comparing cumulative excretion of β-lyase–dependent fluoroacids (3,3,3-trifluoro-2-(fluoromethoxy)propanoic acid plus trifluorolactic acid) and mercapturic acids (alkane plus alkene). The ratio of fluoroacids to mercapturates was approximately 1.5:1 after 1 day (determined to permit comparison with animal data, see below) and 3:1 after 3 days.
Although haloalkene metabolism and toxification by the glutathione-dependent β-lyase pathway has been amply demonstrated in animals, relatively few haloalkenes have been shown to undergo metabolism by this route in humans, and even less quantitative information is available. Urinary excretion of the mercapturic acid metabolites of trichloroethene 39–41and tetrachloroethene 42,43was recently shown in humans receiving occupational or deliberate exposure, demonstrating the biosynthesis of glutathione- and cysteine S- conjugates. Urinary excretion of chloroacetic acid after trichloroethene 41and dichloroacetic acid after tetrachloroethene 43showed that β-lyase–dependent cysteine S- conjugates metabolism can also occur in humans. Compound A appears to be the first fluoroalkene shown to undergo metabolism in humans by the glutathione-dependent β-lyase pathway, the first fluoroalkene in which metabolism by N -acetylation and renal β-lyase has been quantified and may be an excellent probe to explore the toxicologic significance of these pathways in humans.
Limited comparisons of quantitative metabolism of compound A and other haloalkenes in humans are available. After occupational exposure to 200–400 ppm · h tetrachloroethene, mercapturic acid excretion in spot urine samples was 0.010–0.015 nmol/mg creatinine. 42In volunteers exposed to 60, 120, or 240 ppm · h tetrachloroethene, cumulative 35-h mercapturic acid excretion was 45 ± 12, 142 ± 14, and 211 ± 46 nmol, or approximately 0.045, 0.14, and 0.21 nmol/mg creatinine (assuming 1 g/day creatinine excretion), and appeared essentially complete after 35 h. 43After 240, 480, and 960 ppm · h volunteer exposure to trichloroethene, cumulative 48-h excretion of mercapturic acids was 250 ± 40, 370 ± 30, and 430 ± 10 nmol, or approximately 0.25, 0.37, and 0.43 nmol/mg creatinine. 40After trichloroethene intoxication, mercapturic acid excretion on days 1–3 was 0.26, 0.42, and 1.25 nmol/mg creatinine. 41In the present investigation after 78 ± 58 ppm · h compound A, mercapturates excretion was 59 ± 34, 3.5 ± 8.2, and 0 nmol/mg creatinine on days 1–3. Thus mercapturates of compound A were excreted as fast or faster than those of other haloalkenes. Quantitatively, excretion of compound A mercapturates (nmol/kg) also appeared greater than that of other haloalkenes, likely reflecting the fact that these other haloalkenes also undergo considerable cytochrome P450-mediated metabolism, 39–43whereas compound A undergoes comparatively less P450-mediated metabolism. 38We find only one investigation that measured urinary excretion of β-lyase–dependent haloacid metabolites after haloalkene exposure in humans. Völkel et al. found only traces of dichloroacetic acid after 60, 120, or 240 ppm · h tetrachloroethene. 43The present report appears to be the first to quantify β-lyase–dependent haloacid metabolite excretion after haloalkene exposure in humans.
For numerous haloalkenes and their cysteine S -conjugates, 2,3including compound A, 7,18–25metabolism by renal β-lyase to reactive intermediates is considered to confer bioactivation and toxification, whereas N -acetylation is considered to be a detoxication pathway, although other pathways of compound A cysteine S -conjugates metabolism have also been suggested to cause toxicity. 44,45Relative metabolic flux through toxification and detoxication pathways is considered, in part, to determine nephrotoxicity. After inhalation of identical tetrachloroethene concentrations in rats and humans, dichloroacetic acid excretion (reflecting β-lyase–dependent toxification) was orders of magnitude higher in rats than in humans. 43Furthermore, in rats, cumulative dichloroacetic acid excretion was 30- to 80-fold greater than that of the mercapturic acid N -acetyl-S -(trichlorovinyl)-L-cysteine (reflecting detoxication), suggesting considerably greater cysteine S -conjugate metabolism by β-lyase than by N -acetylation. 43In contrast, in humans, β-lyase–dependent metabolism of tetrachloroethene was not observed. 43For tetrachloroethene, greater S -conjugate formation in rats compared with humans, and their more intensive metabolism by β-lyase in rats compared with humans, were interpreted to suggest that reactive intermediate formation in the kidney was greater in rats than in humans after similar exposure conditions, that humans are less sensitive to the nephrotoxic effects of tetrachloroethene, and that risk assessments based on rat data would overestimate human risks. 43
For compound A, similar species differences in metabolism are apparent. Urinary excretion of mercapturic acids and β-lyase–derived fluoroacid metabolites, representing detoxication and toxification, are compared in rats and humans (fig. 7). Rats underwent exposure to 138 ppm · h inhaled compound A, 7the dose which was closest to that in the present patients, and urine metabolite concentrations were determined using the same assays as used herein. 26Rat urine was collected for 24 h, thus 24-h data from the present human investigation are shown for comparison. Fluoroacid excretion (reflecting β-lyase–dependent metabolism) was 15-fold higher in rats than in humans. The ratio of fluoroacids to mercapturates was sixfold greater in rats 10than in humans (1.7). These in vivo results are consistent with previous in vitro data showing that β-lyase–catalyzed metabolism of compound A cysteine-S -conjugates was 8- to 30-times greater in rat compared with human kidneys. 23
In summary, this investigation quantified human exposure and dose of compound A, as well as compound A S -conjugates metabolism by N -acetylation (detoxication) and β-lyase (toxification), and compared them to previous results in rats, which appear more susceptible to compound A nephrotoxicity. Humans, compared with rats, receive markedly lower doses of compound A and metabolize a lower fraction by renal β-lyase than by N -acetylation. These species differences may influence compound A renal effects.
The authors thank Kathy Cox and Dr. Douglas Mautz for their excellent technical assistance.