The modern volatile anesthetics isoflurane, sevoflurane, and desflurane are widely used for general anesthesia and are likely to be administered for the next few decades. Their blood/gas partition coefficients are core pharmacokinetic values because they determine their velocities of uptake into and elimination from the body of a patient.
In clinical practice, recovery from anesthesia may vary when using volatile anesthetics. However, different blood/gas partition coefficients of volatile anesthetics have been published in the past, which may possibly represent specific clinical conditions (Table 1). The obvious range of a specific coefficient may be the result of inhomogeneous study groups or of distinct methodological approaches. For example, healthy volunteers,1–7 small numbers of patients,8–17 patients with specific diseases,4,8,18 or mothers just after delivery of their children19 have been included in the studies for examining blood/gas partition coefficients of adult persons.
Furthermore, in some studies, blood/gas partition coefficients were determined separately in the cohort investigated, and in other studies simultaneously (Table 1).
Against the backdrop of discussions about the “real” blood/gas partition coefficient of a certain volatile anesthetic, we included a clinically relevant patient population with a sufficient sample size. To exclude any interaction, we decided to determine the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane separately in 3 comparable groups in a consistent manner.
The validity of the gas chromatography measurements was checked by determining the distilled water/gas partition coefficients of the 3 volatile anesthetics, comparing them with previously published data.
The possible effects of age, gender, body mass index, hemoglobin concentration, and hematocrit on the blood/gas partition coefficients were studied.
The study was approved by the ethical committee of the Medical Faculty of the University of Leipzig, Leipzig, Germany. We included adult and legally competent patients with ASA physical status I to III. They were scheduled for elective minor and major surgery in the departments of orthopedic surgery, abdominal and vascular surgery, urology, and neurosurgery. Pregnant and breastfeeding women were not included. Written informed consent was obtained from all subjects. We planned to enroll patients for this study within a year.
We recorded the personal data such as age, gender, body mass index as well as the hematocrit and hemoglobin values routinely determined in venous blood after the patient was admitted to our hospital; anonymity was preserved.
Basic single solutions of isoflurane, sevoflurane (both from Abbott GmbH & Co. KG, Wiesbaden, Germany), and desflurane (Pharmacia GmbH, Erlangen, Germany) were prepared as follows: 10 mL of distilled water was placed using a syringe in each of three 21-mL vials.
Ten microliter of the respective fluid anesthetic was transferred into each of the vials. The vials were then closed and shaken until the anesthetics had dissolved completely. Because the concentrations of the anesthetics in the aqueous solutions decreased by taking samples, we regularly added an amount of 10 μL of the anesthetic concerned. Due to the low boiling point of desflurane, the syringe, vial, and anesthetic were stored for 1 hour at 4°C before preparing this particular solution.
Sixteen to 30 mL of blood was collected from each of the fasting subjects right before the induction of anesthesia by retrograde flow through a newly placed IV or arterial catheter. The blood samples were anticoagulated with EDTA. The patients were randomly assigned to 1 of 3 groups for the determination of the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane, respectively.
Seven- to 10-mL aliquots of the blood sample were placed in each of three 21-mL vials (test vials). Fifty microliter of the basic solution of the anesthetic to be studied was added to each test vial which was immediately closed and sealed. The same amounts of the basic anesthetic solution were transferred to three 21-mL vials filled with air (control vials).
For equilibration, the test and control vials were placed in a water bath and shaken vigorously for 15 minutes at 37°C. When this procedure was complete, the concentrations of the respective volatile anesthetic in the gas phases of the test vials and in the control vials were determined by gas chromatography (Agilent Technologies, Lexington, MA). The gas chromatograph was equipped with a 15-m long column (0.32 mm in diameter), and a helium carrier stream flowing at 2 mL/min was directed through the column to a flame ionization detector (at 250°C) which was supplied with hydrogen at 45 mL/min and air at 450 mL/min. The values of the anesthetic concentrations in the gas phases of the test vials and in the control vials were averaged for further calculations of the partition coefficient for that patient’s blood.
In 3 patients of the isoflurane group, 4 patients of the sevoflurane group, and 1 patient of the desflurane group, the blood samples taken before induction of anesthesia were <21 mL, resulting in the anesthetic concentrations being determined in 2 test vials. In 3 cases of the isoflurane group and in 2 cases of the sevoflurane group, the anesthetic concentrations were measured in only 2 control vials for technical reasons.
The blood/gas partition coefficient was then calculated as follows: A control vial contained the total amount of the anesthetic added to each vial. Therefore, we were able to deduce the amount of anesthetic to be found in the blood phase of the test vial by subtracting the averaged anesthetic content in the gas phases of the test vials from that in the control vials. Afterward, it was possible to determine the concentration of the anesthetic in the blood phase and then the blood/gas partition coefficient.
We used the following equation to calculate the blood/gas partition coefficient:
where CC = averaged concentration of volatile anesthetic in the control vials, VC = total volume of the vial (21 mL), CT = averaged concentration of volatile anesthetic in the gas phases of the test vials, and VT = volume of the aliquot of blood transferred to the test vial.
On random study days, we determined the distilled water/gas partition coefficients of isoflurane, sevoflurane, and desflurane, applying the same procedure as described above. Because distilled water is a standardized liquid, our distilled water/gas partition coefficients could be compared directly with the values previously reported.
An a priori power analysis was not conducted. Results are expressed as mean ± SD. Because we did not perform an a priori power analysis and performed >20 statistical comparisons, only P values <0.01 were considered to denote significant differences. For the same reason, the 99% confidence limits of our blood/gas partition coefficients and correlation coefficients (see below) were calculated.
The Kolmogorov-Smirnov test (exact significance) was used to confirm the normality of distribution. By performing the analysis of variance test, we compared age, body mass index, the concentration of hemoglobin, and the hematocrit of the 3 study groups the patients had been randomly assigned to for the determination of the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane, respectively. The distributions of gender, ASA physical status, and treating hospital department among the 3 collectives were compared using the χ2 test. We applied the t test for independent samples to examine whether the blood/gas coefficients differed significantly between male or female patients and between the blood samples that were drawn venously or arterially.
Correlation was assessed determining the Pearson coefficients, r, for which 99% confidence intervals were also calculated.
To compare the values of the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane reported thus far with our results, we calculated the mean for each anesthetic from the previous study results concerned, weighted according to the number of determinations or of patients included. Although the blood/gas partition coefficients generally were determined in vitro, in 2 study groups, they were measured in vivo.8 The results of these were not included in the above-mentioned calculation. We applied the 1-sample t test in the subgroups of isoflurane, sevoflurane, and desflurane, respectively, to confirm the significant differences between our blood/gas partition coefficients and those reported thus far.
Statistical analysis was performed using the SPSS Statistics version 21 software (IBM, Armonk, NY).
We studied 134 patients in the year allocated for enrolling patients into this study. Data for 4 study participants in both the isoflurane and the sevoflurane groups and for 6 subjects in the desflurane group were excluded for technical reasons. Thus, final data analysis was performed on 120 patients. The isoflurane and sevoflurane groups ultimately consisted of 41 study participants each, whereas the desflurane group comprised 38 patients.
Values of blood/gas partition coefficients, age, body mass index, the hemoglobin concentration, and hematocrit were normally distributed for each group (P = 0.057 for age in the isoflurane group, all remaining P > 0.215). Table 2 shows the age, gender, body mass index, ASA physical status, the hemoglobin concentration and hematocrit of the study persons, and the treating hospital department. There were no statistically significant differences among the values of the parameters presented for the 3 groups (all P > 0.183).
Our blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane along with the weighted means of the partition coefficients from previous studies are presented in Table 3. Furthermore, the 1-sample t test confirmed the significant differences between our blood/gas partition coefficients and those reported thus far (all P ≤ 0.001). Mean values of this study are 5.07%, 12.12%, and 7.55% higher for isoflurane, sevoflurane, and desflurane, respectively, than the previously published mean values.
We found no significant correlations of our blood/gas partition coefficients with patient age, body mass index, hemoglobin concentration, or hematocrit (Table 4). There were only trends for small correlations between the blood/gas partition coefficient of isoflurane and hemoglobin concentration (r = 0.32 [99% CI, −0.09 to 0.69]; P = 0.041) and hematocrit (r = 0.37 [99% CI, −0.13 to 0.73]; P = 0.016). Furthermore, there were no significant differences between female and male patients or between the arterial and venous blood samples (Table 3).
There were 24 calculations of the distilled water/gas partition coefficient of isoflurane, with 1 excluded for technical reasons. Twenty-five determinations of the distilled water/gas partition coefficient of sevoflurane and 30 for desflurane were performed (Table 5). Values of water/gas partition coefficients were normally distributed (the 3 P > 0.250). The data gathered confirmed the validity of the gas chromatography method used in this study.
In this study, we determined the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane in a common, clinically relevant patient collective with a sufficient sample size. Furthermore, the possible influences of variables such as age, body mass index, gender, hemoglobin concentration, or hematocrit on those partition coefficients were tested.
Blood/gas partition coefficients of this study are 5.07%, 12.12%, and 7.55% higher for isoflurane, sevoflurane, and desflurane, respectively, than the previously reported mean values. Previous studies included small numbers of study persons8–17 or particular patient groups, for example, healthy persons,1–7 patients experiencing a common kind of disease,4,8,18 or mothers immediately after giving birth.19 Hu et al.4 reported the highest blood/gas partition coefficient of desflurane (0.59 ± 0.05), but this was determined in a subgroup of 20 cardiac surgical patients, which is a selective patient collective.
In most studies (Table 1), the blood/gas partition coefficient of a single volatile anesthetic was measured or those of several anesthetics were determined simultaneously. With regard to isoflurane, dilution with infusion may have led to the determination of the 2 lowest blood/gas coefficients as blood samples were drawn during operation.8
The results of our study may depict the clinical reality to a greater extent because it included a larger, clinically relevant, and adult patient population. Furthermore, we decided to measure the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane separately in 3 comparable groups in vitro, in a consistent manner, to exclude any interaction. As in nearly all studies, blood samples were taken from fasting subjects.
Our blood/gas partition coefficients are higher than the values that have been published previously, in particular, those of sevoflurane and desflurane. Although the blood/gas partition coefficient of isoflurane has been reported in many publications, further studies will provide more information on the range of the coefficients for sevoflurane and desflurane.
The gas chromatography technique we used for determining the partition coefficients differs from those which have been applied previously. To test the precision of our method, we calculated the SDs of the values of the anesthetic concentrations in the gas phases of the test vials and of those in the control vials from their respective mean value. In most cases, the SDs were <2.5% of the mean value: for 35 patient data in the isoflurane group, 39 patient data in the sevoflurane group, and 32 patient data in the desflurane group. The remaining SDs were <3.5%, with 2 exceptions in the desflurane group. These equaled 7.7% of the corresponding mean value. This is seen to be proof of the consistency of our gas chromatography measurement.
To verify the validity of our gas chromatography technique, we determined the distilled water/gas partition coefficients of isoflurane, sevoflurane, and desflurane. Those of isoflurane and sevoflurane are similar to the results1,20–23 based on other techniques. The one of desflurane has not been published thus far. Nevertheless, the fluid/gas partition coefficients of isoflurane, sevoflurane, and desflurane in crystalloid solutions have been shown to decrease when the osmolarity of the solution increases.22–24 Therefore, we believe our distilled water/gas partition coefficient of desflurane to be exact because it is slightly higher than many saline/gas partition coefficients of desflurane published previously (Table 5).5–7,13,24 These findings led us to the conclusion that our method is valid. Similarly, in another study, the saline/gas partition coefficients and the oil/gas partition coefficients of isoflurane, sevoflurane, and desflurane were determined and used to validate a chromatography method by comparing them with previously published values.7
According to our results, age, body mass index, gender, hemoglobin concentration, or hematocrit do not have significant effects on the blood solubility of the 3 volatile anesthetics. We found only trends for small correlations between the blood/gas partition coefficient of isoflurane and hemoglobin and hematocrit. Although the upper 99% confidence limit for r was approximately 0.7 for both, in each case, the lower level was <0. The upper limit of 0.7 does not preclude a modestly large correlation, but the wide range in the confidence interval indicates that far more subjects would need to be studied to better characterize if there is a moderate correlation. There was no significant difference between arterial and venous blood/gas partition coefficients.
The influences of the body mass index and gender of study persons have not been tested before. No significant correlation has been reported between the age of adults and the blood/gas partition coefficient of isoflurane,8,11,16 sevoflurane,12,16 and desflurane,13 which is confirmed by our results.
Previously published effects of the hemoglobin concentration or the hematocrit on the blood/gas partition coefficients of isoflurane, sevoflurane, and desflurane are not homogeneous. The decrease in the hematocrit attributable to dilution of whole blood with saline resulted in a reduction of blood solubility of the volatile anesthetics.6,8,14,18 When blood of single volunteers was separated into its plasma and red cell fractions and afterward different amounts of the respective fractions were recombined, the blood/gas partition coefficient of isoflurane increased as hematocrit decreased.9
On the contrary, there were no significant correlations between the blood/gas partition coefficients of isoflurane, sevoflurane, or desflurane and the hemoglobin concentrations or the hematocrit among individuals.4,8,12,13 Generally, our findings correspond with the results of these studies, but we saw trends for small correlations of isoflurane blood/gas partition coefficient with hemoglobin concentration and hematocrit, which are not significant and probably of minor clinical importance.
In summary, we found the blood/gas partition coefficients to be 5.07%, 12.12%, and 7.55% higher for isoflurane, sevoflurane, and desflurane, respectively, than previously reported and that the coefficients of modern volatile anesthetics remain unaffected by age, gender, body mass index, hemoglobin, or hematocrit.
The extent to which balanced anesthesia may be influenced by these findings cannot be identified, but this is probably of secondary relevance. Modern volatile anesthetics are well-controlled drugs; sevoflurane and desflurane continue to be short-acting volatile anesthetics despite their possibly slightly higher blood/gas partition coefficients. Perhaps, the blood/gas partition coefficient of a certain volatile anesthetic should not be given as an absolute value but instead within a defined range.
Name: Tobias Esper, MD.
Contribution: This author helped design the study. He conducted the study, collected and analyzed the data, and prepared the manuscript.
Attestation: Tobias Esper approved the final manuscript. He attests to the integrity of the original data and the analysis reported in this manuscript. He is the archival author.
Name: Markus Wehner, MD.
Contribution: This author designed the study. He helped conduct the study, collect the data, and analyze the data.
Attestation: Markus Wehner approved the final manuscript.
Name: Claus-Dieter Meinecke, PhD.
Contribution: This author helped collect and analyze the data.
Attestation: Claus-Dieter Meinecke approved the final manuscript.
Name: Henrik Rueffert, MD.
Contribution: This author conducted the study, helped write the manuscript, and approved the final version.
Attestation: Henrik Rueffert approved the final manuscript. He attests to the integrity of the original data and the analysis reported in this manuscript.
This manuscript was handled by: Marcel E. Durieux, MD, PhD.
The authors thank Marita Ziepert, PhD, and York Hilger, Senior Statistical Consultant, for consulting on statistical data. The authors gratefully acknowledge the contribution to the determinations of the blood/gas partition coefficients from Daniela Geier. The authors thank Matthew Rockey for proofreading and Timolaos Rizos, MD, PhD, for comments on the manuscript and fruitful discussions.