Plasma Inorganic Fluoride Concentrations after Sevoflurane Anesthesia in Children

After Institutional Review Board approval at The Hospital for Sick Children, Toronto, and Children's Hospital of Pittsburgh and approval by the federal regulatory agencies in both Canada and the United States, written consent was obtained from the parents of 120 children, ASA physical status 1 or 2 aged 1-12 yr who were scheduled for elective surgery. Because this study was conducted as part of a larger study, the methods outlined later have been reported in part. [11].

Children scheduled for surgery with an anticipated loss of less than 10% of their blood volume and duration of 1-5 h were eligible for participation in the study. The surgical procedures included urologic (not involving the kidneys), general, plastic, orthopedic, or ear, nose, and throat surgery (excluding surgery of the tracheobronchial tree). Children with a history of pulmonary, cardiac, renal, or hepatic disease were excluded as were those with a history or family history of muscle disease (e.g., malignant hyperthermia) or those who had received any experimental drug in the preceding 28 days. Children who received medications known to be nephrotoxic, that increase hepatic enzyme activities or that affect MAC were also excluded from the study.

All children were fasted for approximately 4 h after clear fluids and none were premedicated. They were randomly assigned to one of three treatment groups before induction of anesthesia: group 1 received sevoflurane in air/oxygen 30% (n = 40), group 2 received sevoflurane in N sub 2 O 70%/oxygen 30% (n = 40), and group 3 received halothane in N²O 70%/oxygen 30% (n = 40). Anesthesia was induced by inhalation according to the treatment assignment using a Mapleson D or F anesthetic circuit. The fresh gas flow included nitrous oxide, air, and/or oxygen according to the treatment assignment, at a total fresh gas flow of 3-6 l/min. After tracheal intubation under deep inhalational anesthesia, ventilation was controlled mechanically to maintain the end-tidal carbon dioxide tension between 30 and 40 mmHg.

Anesthesia was maintained with 1.0-1.3 MAC sevoflurane or halothane. [10,12]The MAC calculations, based on published data for sevoflurane and halothane, [10,12]did not include the contribution of nitrous oxide: children 1-2 yr received sevoflurane 2.6% or halothane 1.0% and children 3-12 yr received sevoflurane 2.5% or halothane 0.9%. The end-tidal concentrations of sevoflurane and halothane were decreased to 1.0 MAC for at least 15 min before the termination of anesthesia. End-tidal concentrations of sevoflurane and halothane were monitored continuously using a calibrated Datex Capnomac Ultima (Helsinki, Finland). The monitor was calibrated immediately before administration of each anesthetic. Intravenous lactated Ringer's solution was administered throughout the study to replace any fluid deficit and to provide maintenance fluids.

Immediately after loss of consciousness, an intravenous catheter was inserted into an arm vein to collect blood samples. Blood samples (3 ml) were collected for determination of the plasma [Fluorine sup -] immediately after induction of anesthesia, at the termination of anesthesia, and at 1, 4, 6, 12, and at either 18 or 24 h after discontinuation of anesthesia. Two additional 3-ml blood samples were collected for complete blood count, liver function tests, and the plasma concentrations of electrolytes, urea, and creatinine immediately after induction of and at 18 or 24 h after discontinuation of anesthesia. After these samples were collected in plastic syringes containing small quantities of heparin, they were centrifuged within 1 h of collection. The supernatant plasma was pipetted into plastic containers, sealed, and then stored at -20 degrees C until analysis for [Fluorine sup -].

Plasma inorganic fluoride concentrations [Fluorine sup -] were determined using an Orion ion-specific Ionanalyzer (Orion Research, Boston, MA). The detection limit of the assay was 1.0 micro Meter. The within-batch percent coefficient of variation was 8.6%. The intergroup percent coefficient of variation was 7.9%. The plasma [Fluorine sup -]-time profile for each child was plotted and the area under the plasma [Fluorine sup -]-time curve (AUC⁰→Ct) determined using the trapezoidal rule from time 0 to the last blood sample (C^t). The AUC sub Ct→infinity was calculated using the ratio, C^t/B, where B is the elimination rate constant. The AUC⁰→infinity was the sum of the AUC^sub0→Ct and the AUC^Ct→infinity.

Data are presented as means+/-standard deviation. Regression analysis was used to determine the relationship (expressed as the coefficient of determination (r²)) between the anesthetic exposure (MAC *symbol* h) and both the AUC for Fluorine sup - and the peak plasma [Fluorine sup -], and to determine the relationship between the AUC and the peak [Fluorine sup -]. Parametric data were compared using the unpaired Student's t-test. Statistical significance of P < 0.05 was accepted.

Demographic data of the children in groups I and II within each center were similar. Accordingly, we combined the data from these two groups within each center. The duration of anesthesia for sevoflurane and halothane, respectively, in the two centers was similar. Here too, we combined the data of sevoflurane and halothane from the two centers, respectively.

The mean ages of the children and duration of anesthesia (in MAC *symbol* h) in the sevoflurane and halothane groups were similar (Table 1). The duration of sevoflurane anesthesia exceeded 5 MAC *symbol* h in six children, with the greatest exposure being 8.9 MAC *symbol* h.

The peak recorded plasma [Fluorine sup -] after sevoflurane was eightfold greater than that after halothane (P < 0.0001; Figure 1and Table 1). The peak recorded plasma [Fluorine sup -] occurred 1 h after the termination of anesthesia in 77 of the 80 children. In the remaining three children, the peak value occurred between 4 and 6 h after anesthesia. The range of peak recorded [Fluorine sup -] after sevoflurane was 7-28 micro Meter. The maximum recorded [Fluorine sup -], 28 micro Meter, occurred in a child who had received 7.0 MAC *symbol* h of sevoflurane. In all children who received sevoflurane, the plasma [Fluorine sup -] decreased to < 6.2 micro Meter by 24 h after termination of the anesthetic. The plasma [Fluorine sup -] after halothane was less than 10 micro Meter throughout the entire 24-h study period.

The relationships between peak plasma [Fluorine sup -] and sevoflurane exposure (r²= 0.50; Figure 2) and between AUC⁰→infinity for inorganic fluoride and MAC *symbol* h sevoflurane (r²= 0.57; Figure 3) were linear. The peak plasma [Fluorine sup -] after halothane was independent of the halothane exposure (r²= 0.007; Figure 2). The relationship between the peak plasma [Fluorine sup -] and AUC⁰→infinity for sevoflurane was also linear (r²= 0.66). The AUC⁰→24h after sevoflurane was 84+/-7.5% of the AUC⁰→infinity (Table 1). The AUC⁰→24h for sevoflurane was ninefold greater than the AUC⁰→24h for halothane (Table 1).

The AUC⁰→infinity was determined accurately in 39 of the 80 children who received sevoflurane (Figure 3and Table 1). We could not determine an accurate AUC⁰→infinity in 41 of the children because of incomplete blood collection as a result of inadvertent removal of the intravenous cannula and/or patient or parental refusal to allow for the continued blood sampling in the recovery period. As a result, we could not accurately estimate the terminal elimination half-life of inorganic Fluorine sup - and thus the AUC⁰→infinity for those 41 children.

The mean plasma concentration of urea 24 h after both sevoflurane and halothane anesthesia decreased significantly compared with preanesthetic values, respectively (Table 1). In contrast, plasma creatinine level was unchanged after both anesthetics compared with preanesthetic values. One child developed pulmonary edema 48 h after surgery as a result of fluid overload that occurred during the postoperative period. His serum creatinine clearance increased from 39 mmol *symbol* 1 sup -1 preoperatively to 89 mmol *symbol* 1 sup -1 (normal < 60 mmol *symbol* 1 sup -1) 24 h after 4.8 MAC *symbol* h sevoflurane anesthesia. The serum creatinine concentration returned to baseline concentrations 3 days postoperatively. The peak plasma [Fluorine sup -] in this patient was 25.5 micro Meter. In this case, the investigator deemed the transient increase in plasma creatinine concentration to be unrelated to the administration of sevoflurane.

The introduction of sevoflurane into clinical anesthesia has been clouded by concerns about the potential risk of nephrotoxicity after its use. Two theoretical sources for the nephrotoxicity after sevoflurane are the plasma concentration of inorganic fluoride, an in vivo metabolite of sevoflurane and compound A, an in vitro degradation product of sevoflurane in the presence of soda lime and baralyme. The purpose of this study was to address the possible role of inorganic fluoride in causing nephrotoxicity in children by characterizing the plasma [Fluorine sup -]-time curve after 1-5 MAC *symbol* h sevoflurane. We found that the peak recorded plasma [Fluorine sup -] after 2.7 MAC *symbol* h (range of 1.1-8.9 MAC *symbol* h) sevoflurane in children aged 1-12 yr occurred approximately 1 h after discontinuation of sevoflurane, was 15.8 micro Meter and decreased rapidly to < 6.2 micro Meter in all children by 24 h (Figure 1).

The plasma [Fluorine sup -]-time relationship after inhalational anesthetics is determined by three variables: (1) the duration of anesthesia; (2) the lipid solubility of the anesthetic, and (3) metabolism of the anesthetic. Both the lipid solubility and metabolism of the anesthetic are properties of the anesthetic. The metabolism (and estimated percent metabolized in vivo in the absence of enzyme induction) of the inhalational anesthetics follows the order: methoxyflurane (50%) > halothane (15-20%) > sevoflurane (5%) > enflurane (2.4%) > isoflurane (0.2%) > desflurane (0.02%), [1,13,14]although published evidence suggests that the extent of metabolism of methoxyflurane, halothane, and enflurane may exceed these estimates. [15]All ether anesthetics are defluorinated in vivo by the cytochrome P450 microsomal enzyme system and, quantitatively, the plasma [Fluorine sup -]-time relationship parallels the relative metabolism of these anesthetics for a given duration of anesthesia. In contrast to the ether series of anesthetics however, the alkane anesthetic halothane is metabolized substantially in vivo but releases very little inorganic fluoride (Figure 1and Figure 2). Accordingly, the order of both the peak [Fluorine sup -] and the AUC for [Fluorine sup -] is: methoxyflurane > sevoflurane > enflurane > isoflurane [nearly equal] halothane > desflurane. On the basis of the pharmacokinetic curves, the terminal elimination of Fluorine sup - for sevoflurane is similar to that of enflurane. [6,16].

The pharmacokinetics of Fluorine sup - in children have been investigated after only 2 ether anesthetics, enflurane and methoxyflurane. The peak [Fluorine sup -] after [nearly equal] 1.5 MAC *symbol* h enflurane, 6-10 micro Meter, [17]is similar to that obtained after sevoflurane when the latter was adjusted for the anesthetic dose, whereas the peak [Fluorine sup -] after methoxyflurane, 22 micro Meter, is approximately 50% greater than that obtained after sevoflurane. [18]Thus, the pharmacokinetics of inorganic fluoride after sevoflurane in children are similar to those after enflurane but less than those after methoxyflurane--a relationship similar to that reported in adults. [6,18].

The pharmacokinetics of Fluorine sup - after inhalational anesthesia in children have not been compared directly with those in adults. In the case of methoxyflurane, the peak [Fluorine sup -] in children is approximately half that in adults for a similar duration of anesthesia. [18]We calculated the AUC⁰→48h for Fluorine sup - based on published data [18]and found that the AUC⁰→48h in children was half that in adults for the same MAC *symbol* h. For sevoflurane, the peak [Fluorine sup -] in children is less than half that reported after sevoflurane in adults although the dose of sevoflurane in children was 30% less than that in adults. The AUC⁰→infinity in children, however, was only one fifth that in adults. These data suggest that in the case of both methoxyflurane and sevoflurane, peak [Fluorine sup -] and AUC for Fluorine sup - are less in children than they are in adults for a similar anesthetic exposure. This may be attributed to several differences between children and adults including decreased metabolism and lower solubility of the anesthetic in children, more rapid uptake by Fluorine sup - in tissues (such as bone) and more rapid renal elimination of Fluorine sup - in children than in adults. [18].

None of the children in this study developed clinical or biochemical evidence of renal insufficiency postoperatively as evidenced by polyuria or an increase in plasma creatinine concentration. This is not surprising in view of the low plasma [Fluorine sup -] measured (maximum value of 28 micro Meter). Although the plasma concentrations of urea after both sevoflurane and halothane anesthesia were significantly less than preoperative values (Table 1), this isolated biochemical change is most likely explained by the intraoperative administration of large volumes of intravenous balanced salt solutions. Our assessment of renal function, however, was incomplete because we did not measure urine osmolality, renal concentrating ability, or urinary elimination of renal tubular enzymes. Despite the incomplete testing performed, it is unlikely that nephrotoxicity will occur after sevoflurane in children, because similar studies in adult volunteers and patients even under conditions that are likely to yield increased plasma [Fluorine sup -] (i.e., after prolonged sevoflurane anesthesia (up to 11.5 MAC *symbol* h), renal insufficiency, obesity, and enzyme induction) failed to produce evidence of a renal concentrating defect or nephrotoxicity, [16,19,20]although this has been questioned recently. [21].

A second theoretical cause of nephrotoxicity after sevoflurane anesthesia is compound A, an in vitro degradation product of sevoflurane in soda lime and baralyme. Compound A causes dose-dependent histologic changes in the kidneys of rats, [22,23]although there has been no evidence of renal insufficiency attributable to compound A in humans to date. In the current study, we used Mapleson D and F circuits; we did not use circle circuits and carbon dioxide absorbers. Accordingly, the risk of nephrotoxicity from compound A in the current study was zero.

In conclusion, 2.7 MAC *symbol* h (range: 1.1-8.9 MAC *symbol* h) sevoflurane in children yields peak recorded plasma [Fluorine sup -] that are low, similar to those after enflurane and a plasma [Fluorine sup -]-time curve that has a small AUC compared with adults.

The authors thank Dr. J. Gandolfi and his colleagues at the University of Arizona, Tucson, for the analysis of the plasma concentrations of inorganic fluoride.