Anesth Analg 2003;96:356-362
© 2003 International Anesthesia Research Society
CARDIOVASCULAR ANESTHESIA
Isoflurane and Desflurane Uptake During Liver Resection and Transplantation
Jan F.A. Hendrickx, MD*,
Michael K. Dishart, MD*, and
Andre M. De Wolf, MD*,
*Department of Anesthesiology and CCM, University of Pittsburgh School of Medicine, Pennsylvania; and
The Feinberg School of Medicine, Northwestern University, Chicago, Illinois
Address correspondence and reprint requests to Andre M. De Wolf, MD, Department of Anesthesiology, The Feinberg School of Medicine, Northwestern University, 251 E. Huron St., F5–704, Chicago, IL 60611–2908. Address e-mail to a-dewolf{at}nwu.edu
 |
Abstract
|
|---|
When reducing fresh gas flows, the course of the vaporizer dial settings required to maintain a constant end-expired concentration of a potent inhaled anesthetic becomes more dependent on the uptake pattern of the inhaled anesthetic. However, the uptake pattern of potent inhaled anesthetics during prolonged procedures remains poorly quantified. Therefore, we determined isoflurane and desflurane uptake (Viso and Vdes, respectively) during liver resection (LR, n = 17) and orthotopic liver transplantation (OLT, n = 18) using a liquid injection closed-circuit anesthesia technique maintaining the end-expired concentration at 0.8% and 4.5%, respectively. Individual and average uptake curves were fit to a series of mathematical functions and compared with the square root of time and four-compartment models. Cumulative doses of isoflurane and desflurane after 1 and 3 h in the LR group and after 1, 3, and 8 h in the OLT group were correlated with demographic variables and each patient’s average cardiac output and cardiac index. Average uptake was best described by a biexponential fit: Viso (LR) = 1.5 x (1 - e-t x 0.525) + 16.4 x (1 - e-t x 0.00506) (R2 = 0.9996); Viso (OLT) = 1.4 + 3.1 x (1 - e-t x 0.472) + 26.7 x (1 - e-t x 0.00307) (R2 = 0.9994); Vdes (LR) = 2.7 x (1 - e-t x 0.763) + 28.7 x (1 - e-t x 0.00568) (R2 = 0.9984); and Vdes (OLT) = 1.4 x (1 - e-t x 0.472) + 26.7 x (1 - e-t x 0.00307) (R2 = 0.9994). Uptake showed significant interindividual variability, and correlations between uptake variables and patient characteristics were inconsistent. The rate of uptake decreased more slowly then predicted by the uptake models. Because neither existing models nor patient characteristics accurately predict uptake in the individual patient, anesthesia techniques involving the use of low fresh gas flows will continue to have to rely on drug monitoring. However, the slowly decreasing rate of uptake during prolonged procedures suggests that the number of vaporizer adjustments to keep the end-expired concentration constant should be limited.
IMPLICATIONS:The uptake of isoflurane and desflurane during prolonged surgery did not consistently correlate with cardiac output and patient characteristics and differed from two frequently used uptake models. The slowly decreasing rate of uptake implies that the number of vaporizer adjustments during prolonged low-flow anesthesia should be limited.
|
Introduction
|
|---|
Physiological models like the square root of time (SqRT) model and the four-compartment (4C) model (recently modified to a five-compartment model) describe uptake of inhaled anesthetics as the sum of uptake by the individual organs, which can theoretically be predicted based on organ blood flow, organ volume, and organ partition coefficient of the volatile anesthetic (1–3). Both models have been used to simulate uptake kinetics of inhaled anesthetics and to predict the required vaporizer dial settings (1,2). It is therefore important to clinically validate these models to ensure that predictions based on these models produce meaningful results (4). The study of the uptake pattern of inhaled anesthetics is particularly important to the practice of low-flow and closed-circuit anesthesia because the more rebreathing occurs, the more the uptake pattern will be reflected in the vaporizer dial setting to maintain a constant end-expired concentration (1). The rate of desflurane, isoflurane, and sevoflurane uptake (mL/min) in adult patients was found to decrease significantly less than predicted by these models (5–7). A changing cardiac output (CO) could have accounted for these discrepancies (8), but sevoflurane uptake was still found to differ from that predicted by the SqRT and 4C models when measured CO was included in the formulas (9). However, because these studies were limited to 1 h, and because pharmacokinetics should not be extrapolated beyond the domain of the original research (10), we determined isoflurane and desflurane uptake (Viso and Vdes, respectively) during prolonged closed-circuit anesthesia in patients undergoing liver resection (LR) or orthotopic liver transplantation (OLT). We also examined whether there were correlations between cumulative doses of the inhaled anesthetics after 1, 3, and 8 h, and patient characteristics, CO, and cardiac index (CI), and we compared the uptake pattern with the SqRT model and the 4C model.
|
Methods
|
|---|
The hospital’s Biomedical IRB approved the study, and written informed consent was obtained from all patients. Thirty-five adult ASA physical status I–IV patients of both sexes undergoing LR (n = 17) or OLT (n = 18) were enrolled.
The patients’ age, sex, height, and weight were recorded, and body surface area (BSA) was calculated. After breathing oxygen for a few minutes (end-expired nitrogen concentration <3%), anesthesia was induced with thiopental (3–4 mg/kg) and fentanyl (1–3 µg/kg) IV. Intubation of the trachea was facilitated by succinylcholine (1 mg/kg). The circle system was closed after tracheal intubation, and liquid isoflurane or desflurane were injected into the inspiratory limb of the circle system using a syringe pump with a volumetric accuracy of ±2%. To prevent inadvertent vaporization or boiling (desflurane) at room temperature, an ice bag was placed on the syringes containing the liquid anesthetic. Rapid vaporization of isoflurane or desflurane in the circuit was ensured by heating the metal injection port using warm air and verified by visual inspection. After the start of the isoflurane or desflurane infusions, an end-expired isoflurane or desflurane concentration of 0.8% or 4.5%, respectively, was obtained as rapidly as possible and maintained throughout the procedure by adjusting the infusion rate. Oxygen flow was titrated to keep the volume of the bellows at end-expiration constant at approximately 2 cm below the top of the bellows housing.
Ventilation was controlled, and normocapnia was maintained throughout the procedure. Anesthetic gas concentrations were monitored with an anesthetic analyzer (Rascal II Anesthetic Gas Monitor, Ohmeda, Salt Lake City, UT). To avoid N2 accumulation secondary to air entrainment by this monitor and to avoid errors introduced by having to repeatedly flush the anesthesia circuit to washout N2, O2 was used as the carrier gas, and sampled gases (200 mL/min) were not redirected into the circuit. Only two different Ohmeda Plus anesthesia machines were used to study all the patients, with Baralyme as the absorber. The combined volume of the anesthesia circuit and an estimated functional residual capacity of 2 L was 9 L. When ventilating a test lung at peak inspiratory pressures of 20 cm H2O with the anesthetic analyzer connected at the Y-piece, 0.6 mL of liquid isoflurane or 2.1 mL of liquid desflurane had to be injected to obtain 0.8% or 4.5%, respectively; the liquid infusion rates to maintain these concentrations during 3 different 5-h-long test runs were consistently 0.55 and 0.65 mL/h of isoflurane and 3.9 and 4.5 mL/h of desflurane with each respective anesthesia machine. All the individual uptake data have been corrected by this amount of anesthetic lost through gas sampling and leakage and possible absorption by circuit components. An additional correction was applied for blood loss: at 0.8% isoflurane and 4.5% desflurane, 10 L of blood contains 0.57 mL of liquid isoflurane and 0.9 mL of liquid desflurane, respectively.
After placement of a flow-directed pulmonary artery catheter, CO was measured every 20 min by thermodilution. For each patient, mean CO and CI after 1 and 3 h (LR groups) and after 1, 3, and 8 h (OLT groups) were calculated by averaging all the CO and CI obtained for that patient over those time periods.
The cumulative dose of liquid anesthetic injected over time and end-expired anesthetic concentrations were recorded every minute during the first hour and every 5 min thereafter.
Individual and average uptake curves (including the initial bolus to the circuit and functional residual capacity) were fit to a series of mathematical functions (nonlinear regression), including, but not limited to, mono-, bi-, and triexponential functions, with and without y-intercept (Table Curve 2D Automated Curve Fitting Software, Jandel, San Rafael, CA; GraphPad Prism, GraphPad Software, Inc, San Diego, CA). Whereas a large number of equations mathematically accurately (R2 > 0.999) described the uptake curves, exponential functions were chosen because of their simplicity and because exponential functions are used frequently to describe biological processes (tissues ultimately have to saturate, which is an exponential process). The goodness of fit was determined by R2 values, residuals, sum of squares, number of runs, confidence limits of the variables, and visual inspection. Correlations of the cumulative doses after 1 and 3 h (LR groups; Viso 1 h, Viso 3 h, Vdes 1 h, and Vdes 3 h) and after 1, 3, and 8 h (OLT groups; Viso 1 h, Viso 3 h, Viso 8 h, Vdes 1 h, Vdes 3 h, and Vdes 8 h) with age, height, weight, BSA, and with each patient’s average CO and CI after 1, 3, or 8 h were evaluated using linear regression analysis. Finally, individual and average uptake patterns were visually compared with the uptake predicted by the SqRT model and the 4C model using measured CO in their formulas. The formulas used to calculate uptake according to the SqRT and 4C models can be obtained from the authors. Values are presented as mean ± SD.
|
Results
|
|---|
The end-expired isoflurane and desflurane concentrations were 0.8% and 4.5%, respectively, within 3–5 min in all patients. Patient demographic variables, CO, CI, and cumulative doses are presented in Table 1. There was significant interindividual variation in uptake at all times (Figs. 1–4; Table 1). Over the entire duration of the study, the coefficient of variation of Viso and Vdes ranged from 10% to 32%.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. Observed isoflurane uptake (Viso) in the liver resection (LR) group and calculated Viso according to the square root of time model (SqRT model) and four-compartment model (4C model) using measured cardiac output (CO) in the formulas of these models.
|
|
View larger version (27K):
[in this window]
[in a new window]
|
Figure 2. Observed isoflurane uptake (Viso) in the orthotopic liver transplantation (OLT) group and calculated Viso according to the square root of time model (SqRT model) and four-compartment model (4C model) using measured cardiac output (CO) in the formulas of these models.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Figure 3. Observed desflurane uptake (Vdes) in the liver resection (LR) group and calculated Vdes according to the square root of time model (SqRT model) and four-compartment model (4C model) using measured cardiac output (CO) in the formulas of these models.
|
|
View larger version (29K):
[in this window]
[in a new window]
|
Figure 4. Observed desflurane uptake (Vdes) in the orthotopic liver transplantation (OLT) group and calculated Vdes according to the square root of time model (SqRT model) and four-compartment model (4C model) using measured cardiac output (CO) in the formulas of these models.
|
|
Average uptake in all groups was best described by a biexponential fit (Table 2). A monoexponential curve fit with y-intercept described individual Viso with R2 > 0.99 in all but one patient. In 15 of 35 patients, a biexponential function with or without a y-intercept provided a better fit. A biexponential function could not be fit to the other data.
Overall, a higher CO and CI were associated with a larger cumulative dose, except in the desflurane LR group (Table 3). There were inconsistent correlations between cumulative doses and weight and BSA, but correlations between cumulative doses and age and height were poor (Table 3).
View this table:
[in this window]
[in a new window]
|
Table 3. Coefficients of Determination (r2 values) Between Isoflurane and Desflurane Cumulative Dose (Viso and Vdes) and Patient Characteristics, Cardiac Output, and Cardiac Index
|
|
In most individual patients and the groups as a whole, the SqRT model initially overestimated Viso and Vdes but underestimated Viso and Vdes thereafter (Figs. 1–4). The 4C model mostly underestimated Viso and Vdes, especially in the desflurane groups.
|
Discussion
|
|---|
In this study, average Viso and Vdes for up to three hours in patients undergoing LR or up to eight hours in patients undergoing OLT was best described by a biexponential curve. Individual uptake was best described by a monoexponential and, inconsistently, a biexponential curve fit, with or without a y-intercept. Uptake showed significant interindividual variability. Overall, uptake variables correlated well with CO and CI but not with the other patient characteristics. The uptake patterns differed from those predicted by the SqRT model and the 4C model, even when measured CO data were used in the formulas of these models.
The uptake pattern of isoflurane, desflurane, and sevoflurane was recently found to differ from the classical models (SqRT model and the 4C model) in procedures up to one hour (5–7,9). According to our current data, the rate of Viso and Vdes continues to decrease less than previously assumed by these physiological models for up to eight hours. Similarly, Walker et al.(7) found that Vdes remained almost constant after the initial 15–30 minutes of a 3 hour study; once a stable end-expired concentration (8%) was obtained, the rate of injection showed a biexponential decay (y = 10.3 x e-0.77 x t + 0.38 x e-0.094 x t), with a triexponential model providing no better fit. The only other study in which Viso has been measured for prolonged periods unfortunately did not account for leaks around laryngeal masks and tracheostomies and leaks of the breathing system itself (11). Similar to the findings in most other studies, we found inconsistent correlations between patient characteristics and uptake variables (5,6,12,13). The best correlations were found between CO and uptake variables, confirming our previous findings with sevoflurane (9). Even though patient size also has to be a determinant of uptake, the effect of these variables may be minimal over the range of weights and heights studied. In addition, most patients may differ from the ideally compartmentalized 70-kg Mapleson model (14).
Lowe and Ernst (1) apparently used the combination of tissue solubilities and methoxyflurane uptake data to construct their uptake model, the SqRT model, which is conceptually similar to Severinghaus’ (15) uptake model for N2O. Given the lack of accurate on-line anesthetic monitoring at the time, it is remarkable that their predictions regarding uptake of anesthetics are so close to our observations. The discrepancy between the 4C model and our data is less for isoflurane than for desflurane (Figs. 1–4). This may be explained by the exclusion of the vessel poor group (VPG) by the 4C model for desflurane because VPG solubility of desflurane has never been published. We calculated that the VPG uptake of isoflurane is up to 18.4% of total uptake after eight hours.
A changing CO could have affected the uptake pattern: CO increased an average of 14% from the first to the eighth hour, which has been observed previously (16). In patients undergoing OLT however, the CO may have been much lower during the anhepatic stage. In addition, vapor may be lost by prolonged and extensive intestinal exposure. Intestinal losses in animal studies were found to be minimal and therefore unlikely to affect uptake, but good human data are lacking (17).
The uptake pattern itself has relevance to the clinical practice of anesthesia because it is the most important factor of the general anesthetic equation (1). According to this concept, the vaporizer dial setting to attain and maintain a constant end-expired concentration with any fresh gas flow can be predicted if the following are known: (a) uptake of O2, N2O, and the potent inhaled anesthetic, (b) H2O and CO2 production, (c) dead space and alveolar ventilation, and (d) leaks of the system (1). Although the models are not accurate enough to predict drug uptake in the individual patient because of biological variability and therefore cannot predict vaporizer dial settings in the individual patient, drug analysis does allow us to clinically use these uptake models to guide administration of these anesthetics in low-flow situations by letting us make adjustments to compensate for that biological variability in uptake. Thus, although models may only give a rough estimate of the required vaporizer settings at low fresh gas flows, drug analysis allows the clinician to fine-tune the administration of the anesthetic (=vaporizer setting) to obtain the desired alveolar concentration. In addition, drug analysis has allowed the development of new simplified administration schedules during low-flow anesthesia that are based on our previously described uptake patterns during the first hour of the anesthetic (18,19). The long time constants found in the current study suggest that, for clinical purposes, the rate of anesthetic uptake will only slowly decrease during the first two hours after an initial 15-minute wash-in period. This implies that during this time interval, vaporizer settings do not need to be changed according to a clinically more cumbersome to apply mathematical model such as the SqRT model. Other authors have corroborated this (20,21), and our own preliminary results evaluating the general anesthetic equation for isoflurane, sevoflurane, and desflurane for procedures up to one hour also confirm this (22).
In conclusion, uptake patterns of isoflurane and desflurane during prolonged procedures do not consistently correlate with CO and patient demographic variables and cannot be predicted by existing uptake models. Therefore, anesthesia techniques with low fresh gas flows, and therefore with more rebreathing, will continue to have to rely on anesthetic monitoring. Because the rate of uptake decreases more slowly than previously assumed, the number of vaporizer adjustments to keep the end-expired concentration constant should be limited, even when low fresh gas flows are used.
|
Acknowledgments
|
|---|
This work was sponsored by an educational grant from Ohmeda.
The authors thank Michael J. Avram, PhD, Associate Professor of Anesthesiology, Department of Anesthesiology, Northwestern University Medical School, Chicago, IL, for reviewing the manuscript.
|
Footnotes
|
|---|
Presented, in part, at the IARS Meeting, San Francisco, CA, March 14-18, 1997, and the ASA Annual Meeting, San Diego, CA, October 18-22, 1997.
This work was performed at Presbyterian University Hospital, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.
|
References
|
|---|
- Lowe HJ, Ernst EA. The quantitative practice of anesthesia: use of closed circuit. Baltimore, MD: Williams & Wilkins, 1981: 53–97.
- Eger EI II. Uptake of inhaled anesthetics: the alveolar to inspired anesthetic difference. In: Eger EI II, ed. Anesthetic uptake and action. Baltimore, MD: Williams & Wilkins, 1974:77–96.
- Yasuda N, Lockhart SH, Eger EI II, et al. Kinetics of desflurane, isoflurane, and halothane in humans. Anesthesiology 1991; 74: 489–98.[Web of Science][Medline]
- Schnider TW, Shafer SL. Evolving clinically useful predictors of recovery from intravenous anesthetics (editorial). Anesthesiology 1995; 83: 902–5.[Web of Science][Medline]
- Lockwood GG, Chakrabarti MK, Whitwam JG. The uptake of isoflurane during anaesthesia. Anaesthesia 1993; 48: 748–52.[Web of Science][Medline]
- Hendrickx JFA, Soetens M, Van der Donck A, et al. Uptake of desflurane and isoflurane during closed-circuit anesthesia with spontaneous and controlled mechanical ventilation. Anesth Analg 1997; 84: 413–8.[Abstract]
- Walker TJ, Chakrabarti MK, Lockwood GG. Uptake of desflurane during anaesthesia. Anaesthesia 1996; 51: 33–6.[Web of Science][Medline]
- Eger EI II . Complexities overlooked: things may not be what they seem (editorial). Anesth Analg 1997; 84: 239–40.[Web of Science][Medline]
- Hendrickx JFA, Van Zundert AAJ, De Wolf AM. Sevoflurane pharmacokinetics: effect of cardiac output. Br J Anaesth 1998; 81: 495–501.[Abstract/Free Full Text]
- Bouillon T, Shafer SL. Hot air or full steam ahead: an empirical pharmacokinetic model of potent inhalational agents (editorial). Br J Anaesth 2000; 84: 429–31.[Free Full Text]
- Pal SK, Lockwood GG, White DC. Uptake of isoflurane during prolonged clinical anaesthesia. Br J Anaesth 2001; 86: 645–9.[Abstract/Free Full Text]
- O’Callaghan AC, Hawes DW, Ross JAS, et al. Uptake of isoflurane during clinical anaesthesia: servo-control of liquid anaesthetic injection into a closed-circuit breathing system. Br J Anaesth 1983; 55: 1061–4.[Abstract/Free Full Text]
- Westenskow DR, Jordan WS, Hayes JK. Uptake of enflurane: a study of the variability between patients. Br J Anaesth 1983; 55: 595–601.[Abstract/Free Full Text]
- Davis NR, Mapleson WW. Structure and quantification of a physiological model of the distribution of injected agents and inhaled anaesthetics. Br J Anaesth 1981; 53: 399–405.[Abstract/Free Full Text]
- Severinghaus JW. The rate of uptake of nitrous oxide in man. J Clin Invest 1954; 33: 1183–9.
- Price HL, Skovsted P, Pauca AL, et al. Evidence for beta-receptor activation produced by halothane in normal man. Anesthesiology 1970; 32: 389–95.[Web of Science][Medline]
- Laster MJ, Taheri S, Eger EI II, et al. Visceral losses of desflurane, isoflurane, and halothane in swine. Anesth Analg 1991; 73: 209–12.[Abstract/Free Full Text]
- Lerou JG, Verheijen R, Booij LH. Model-based administration of inhalation anaesthesia. 4. Applying the system model. Br J Anaesth 2002; 88: 175–83.[Abstract/Free Full Text]
- Hendrickx JFA, Vandeput DM, De Geyndt AM, et al. Maintaining sevoflurane anesthesia during low-flow anesthesia using a single vaporizer setting change after overpressure induction. J Clin Anesth 2000; 12: 303–7.[Web of Science][Medline]
- Lin CY. Uptake of anaesthetic gases and vapours. Anaesth Intensive Care 1994; 22: 363–73.[Web of Science][Medline]
- Lin CY, Mostert JW, Benson DW. Closed circle systems: a new direction in the practice of anesthesia. Acta Anaesthesiol Scand 1980; 24: 354–61.[Web of Science][Medline]
- Hendrickx JFA, De Ridder K, De Geyndt A, et al. Clinical evaluation of the general anesthetic equation for desflurane [abstract]. Anesth Analg 1999; 88: S341.
Accepted for publication October 9, 2002.
|
Top
Abstract
Introduction
