ManTech Environmental Technology, Inc., P.O. Box 31009, Dayton, Ohio 45437
Received September 20, 2000; accepted November 16, 2000
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: PBPK model; canine; halogenated hydrocarbons; halothane; isoflurane; desflurane; CFC-11.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Determining suitable permissible human exposure limits for halon replacement chemicals necessitates attention to both the level and duration of exposure. A procedure employing physiologically based pharmacokinetic (PBPK) modeling to establish egress times for humans in a halon replacement chemical environment was proposed by Vinegar and Jepson (1996). Implementation of the procedure required modification of the respiratory tract (RT) compartment description commonly used in PBPK models (Vinegar et al., 1998). Usually, the PBPK RT compartment uses a steady-state approximation that results in an initial, unrealistically high arterial blood concentration. The steady-state approximation is suitable for describing inhalation exposures lasting from minutes to hours but is not able to describe the first second to minute of an exposure. Using the model with the modified RT, Vinegar et al. (2000) described a procedure for setting safe acute exposure limits for halon replacement chemicals. They linked the cardiac sensitization endpoint, after 5 min of exposure to the LOAEL concentration, to the associated arterial concentration. This arterial concentration measured in dogs was then used as the target concentration for simulated human exposures. For a given exposure concentration, the time until a human reached the target arterial concentration was considered safe from the potential for cardiac sensitization. However, pharmacokinetic studies in dogs are expensive. If a canine PBPK model could be used to predict the arterial concentration, the major experimental cost would be for the determination of partition coefficients.
The purpose of this study was to modify and validate an existing PBPK model for use with dogs. The model was originally developed to predict blood and exhaled breath chemical concentrations in humans during inhalation exposures of less than 5 minutes duration and includes a breath-by-breath description of inhalation in humans (Vinegar et al., 2000). Several data sets were found in the literature, containing pharmacokinetic data in dogs that were exposed by inhalation to the volatile anesthetics halothane (2-bromo-2-chloro-1,1,1-trifluoroethane), isoflurane (2-chloro-2-[difluoromethoxy]-1,1,1-trifluoroethane), and desflurane (2fluoro-2-[difluromethoxy]-1,1,1-trifluoroethane) and to the chlorofluorocarbon CFC-11 (trichlorofluoromethane). An additional data set was found where CFC-11 was administered by intravenous injection. These data were used to validate the use of the modified model for the dog. We discuss the importance of data quality and appropriateness of parameter values.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Azar et al. (1973) exposed awake male beagles (1114 kg) to CFC-11 via a facemask. Arterial and venous concentrations were presented from 10-min exposures to 0.1%, 0.5%, and 1.0%.
Chiou and Niazi (1973) gave a 3-min intravenous infusion of 28 mg of CFC-11 dissolved in 36 ml normal saline to a 45-kg mongrel dog. Venous concentrations were presented over a 4-h period.
Thomas (1975) exposed anesthetized, artificially ventilated beagles to isoflurane through a tracheal cannula. The cannula was connected to a nonrebreathing valve in a closed loop system that contained a CO2 scrubber and an inflow of O2. Data presented for an 8.5-kg male and a 7.0-kg female dog included total ventilation, respiratory rate, inhaled concentration, exhaled concentration, and arterial concentration.
Frei et al. (1988) exposed anesthetized, artificially ventilated mongrel dogs (2439 kg) separately to isoflurane and halothane through a tracheal cannula. The anesthetics were delivered using a feedback-controlled delivery system that maintained a constant end alveolar concentration of agent. Breathing rate was maintained at 12 breaths per min with tidal volume being varied to maintain arterial carbon dioxide at 40 mm Hg. Data presented included inspired and end-expired concentration, arterial and mixed venous concentration, and cardiac output.
Hughes et al. (1980) exposed anesthetized, artificially ventilated greyhounds (2030 kg) to halothane through a tracheal cannula. Data presented included arterial concentration and cardiac output.
Wilhelm (personal communication) exposed anesthetized, artificially ventilated male beagles (12.8 kg mean weight) for 4 h, separately, to 1.76% isoflurane and 8.94% desflurane through a tracheal cannula. Exhaled concentration data were presented.
A breath-by-breath PBPK model (Vinegar et al., 1998) that was used to simulate short-term (0 to 5 min) human inhalation exposure to various halogenated hydrocarbons was modified to include appropriate parameter values for dogs. The respiratory tract description differs from usual PBPK models in that it accounts explicitly for dead space (upper respiratory tract and tracheobronchial region) and pulmonary region volumes, inhalation-exhalation, the time delay for air and gases to travel through the dead-space volume and the pulmonary region, the absorption of the inhaled gas by pulmonary region tissues, and the exchange of the gas with blood. These enhancements allowed the successful simulation of the first several breaths of exposure, which was not possible with the traditional PBPK description of lung uptake.
Partition coefficients were obtained from the literature and are listed with citations in Table 1. The metabolic constants for halothane were taken from Williams et al. (1996). Vmaxc (maximum metabolic rate) for isoflurane was set to one hundredth and for desflurane to one thousandth of that for halothane (Holaday, 1977
; Holaday et al., 1975
; Sutton et al., 1991
).
|
|
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The data used to validate the breath-by-breath dog model were collected under varying protocols, which resulted in variability in ventilation and cardiac output. The degree of reporting of the values of these parameters also varied amongst the papers. Dogs inhaled CFC-11 while awake and spontaneously breathing (Azar et al., 1973), and were given intravenous infusion while awake and spontaneously breathing (Chiou and Niazi, 1973
). Exposure to the anesthetics isoflurane (Thomas, 1975
), isoflurane and desflurane (Wilhelm, personal communication), isoflurane and halothane (Frei et al., 1988
), and halothane (Hughes et al., 1980
) was accomplished using mechanical ventilation.
Hughes et al. (1980) showed that cardiac output decreases by over 50% during the course of stepped increases in inhaled halothane concentration. Cardiac output decreased from a control level of 0.155 l/min/kg to 0.125, 0.122, 0.084, and 0.066 at 0.5, 1.0, 1.5, and 2.0% halothane, respectively.
Ahlgren et al. (1978) exposed dogs to 1.5% halothane for either 30 or 120 min. Six dogs with average weight of 18 kg had a reduction in cardiac output from 2833 to 1630 ml/min after a 30-min exposure to 1.5% halothane. A second group of 7 dogs with average weight of 15 kg showed cardiac output dropping from 1827 to 1490 ml/min after 120 min of exposure. The 42% drop in cardiac output compared with control level, after a 30-min exposure, was comparable to the 45% drop in cardiac output seen by Hughes et al. (1980), also after 30 min of 1.5% halothane exposure.
Dogs exposed to desflurane while spontaneously breathing showed cardiac outputs of 128, 115, and 96 ml/kg/min and respiratory rates of 37, 47, and 22 at end tidal exposures of 10.3, 12.9, and 15.6%, respectively (Clarke et al., 1996).
Measurements of respiration rate and tidal volume were made in spontaneously breathing dogs exposed for 40 min, via tracheal intubation, to 1.35% halothane or 1.95% isoflurane (Hellebrekers, 1986). They had previously been exposed to up to 4.5% of each agent by face mask in order to anesthetize them for the tracheal intubation. Respiratory rates increased for both agents during the 40-min period, although the rates for halothane were generally nearly twice those for isoflurane. Larger changes were seen in tidal volume, which initially dropped for 5 min, leveled off, and then started to increase at 30 min with values approaching control levels at 40 min. Halothane produced greater decreases in tidal volume than did isoflurane.
In light of the variability of ventilation and cardiac output during the exposures described, it is useful to consider the difference in the behavior of the model, comparing default values for ventilation and cardiac output with values that take into account the changes that might result from the exposure. Of the data sets used, the most complete is that of Frei et al. (1988) (Fig. 4). Measurements of inspired, end-expired, arterial, and mixed venous concentrations were made and reported for 160 min of exposure and 50 min of washout of the anesthetics halothane and isoflurane. Cardiac output was also measured and reported over the same time period. Ventilation was maintained at 12 breaths per min, but tidal volume was varied to maintain constant arterial PaCO2. Periods of hyperventilation and hypercirculation were induced during the course of the experiment. The data provided in this study were useful for probing the effects of ventilation, perfusion, and ventilation:perfusion ratio (VPR) on the ability of the model to simulate a given data set. Since cardiac output data were available, the actual cardiac output was used in the model and ventilation was calculated based on the VPR. The actual arterial, venous, and alveolar concentrations do not show significant change during the course of the exposure, because the anesthetic was administered with a feedback-controlled delivery system. However, initial simulations of arterial, venous, and alveolar concentration all show a sharp rise at around 100 min, which is when hypercirculation was induced with epinephrine. These simulations were performed initially with the assumption that the VPR was 1 throughout the experiment. However, the induced hypercirculation does not necessarily result in a corresponding change in ventilation. Epinephrine has a profound effect on cardiac output but also results in increased ventilation. The increased ventilation would lead to reduced pCO2, and according to the protocol, would have resulted in the tidal volume delivered by the ventilator being reduced to allow a return to normal pCO2 levels. Setting the VPR to 0.5 during the period of hypercirculation resulted in the simulations more closely predicting the actual data (see Fig. 4
, lower curves). Ventilation was measured at times of sample taking for analysis, but the data were not reported. Ventilation and cardiac output are major contributors to the model's ability to simulate actual data, but another contributing factor may be the concomitant changes in the distribution of cardiac output as a result of anesthetic exposure and of other manipulations, such as the intentional induction of hypercirculation reported in this study. Ahlgren et al. (1978) presented data demonstrating effects of halothane exposure on the distribution of cardiac output. Significant changes were seen in flows to some organs. Data were presented for 30 and 60 min of exposure to 1.5% halothane. Unfortunately, there are not enough data to incorporate into the model.
Although Wilhelm (personal communication) measured inspired and expired (end tidal and mixed) agent concentrations, respiratory rate, and tidal volume on a breath-by-breath basis, the only data provided were the expired concentrations at selected time points and the target inhaled concentration. However, the volume of anesthetic administered was also provided. This was used to set the ventilatory parameters for the model. Fixing the frequency at 12 breaths per minute, a tidal volume of 185 ml provided simulated inspired volumes for isoflurane and desflurane of 9377 and 47,632 ml, respectively. The respective measured, inspired volumes were 9388 and 47,661 ml. Using the ventilatory parameters determined in this way resulted in the simulations of alveolar concentration shown in Figure 6.
Hughes et al. (1980) exposed dogs to stepped increases in halothane concentration and measured cardiac output at the end of each half-hour exposure period. Cardiac output data were used in the model so that values for cardiac output were interpolated for times between the half-hour measurement points. The simulation uses the measured cardiac output as input in the model, and ventilation is derived based on a VPR of 1 (Fig. 5).
Development of a PBPK model is an iterative process involving model formulation, collection of pharmacokinetic data to test the model's structure, and often reformulation to account for any shortcoming of the model in simulating the collected data. Ideally, data collection should occur in conjunction with model development, with the modeler having considerable oversight on the quality and quantity of data collected. Often, the data requirements may go beyond pharmacokinetic samples but may include key physiological variables such as ventilation and cardiac output. There are times when it is necessary to develop and validate a model with data that already exist in the literature. This was the case in this study, where the model was already validated for use in humans but its use in dogs required obtaining results of published pharmacokinetic studies. At first glance, this might not appear to be a bad thing. After all, one now has the opportunity to pick from a variety of studies, choose several, and demonstrate the successful simulation of the pharmacokinetic data presented. However, as a modeler, one quickly learns that not all data sets are created equal. Alas, experiments and data resulting from them are obtained to suit the requirements of the original author and not those of the modeler, who may come along later and want to borrow the data for model validation purposes. When, there are a number of papers to pick from, they generally fall into two broad categories: those from which the data, when modeled, are at least in general agreement with model prediction and those from which the data and model are in hopeless disagreement. Focus will remain on the first category where general agreement occurs between model and data. Then focus must turn to the details of the experiment. Were the animals awake or anesthetized? If awake, were they active (excited) or resting? If they were anesthetized, were they artificially ventilated or allowed to breathe spontaneously? If artificially ventilated, what were the frequency and tidal volume settings of the ventilator? In actuality, in any one of these instances, frequency and tidal volume are important parameter values for the model. In many studies involving anesthetics, the respiratory frequency is often set (and reported in the paper) but tidal volume is varied (and not reported in the paper) to maintain pCO2 within acceptable physiological limits. Published studies with anesthetics often report in the methods that ventilation and inspired and expired anesthetic concentrations were measured on a breath-by-breath basis. Data such as these would of course be ideal for validation of a breath-by-breath PBPK model. However, these measurements are generally not the focus of the study and hence do not appear in the paper. One is then forced to make use of bits and pieces from various papers to determine what the range of acceptable values of ventilation and cardiac output could be under circumstances of the described experiment. Model predictions that may have been off by 30% or more often come into line with the data when appropriate adjustments are made to ventilation and cardiac output.
The model did a reasonable job of simulating data obtained from a variety of experimental protocols. It is apparent that the absolute and relative values of ventilation and cardiac output, both of which vary as a result of exposure to anaesthetic agents, are important contributors to the degree of success of a simulation. With an understanding of how these two parameters vary, Monte Carlo simulation techniques could then be used to predict canine blood concentrations during exposure to selected chemicals. These values would then be the target blood concentrations needed for the human risk assessment process for determining safe exposure to Halon replacement chemicals (Vinegar et al., 2000).
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Azar, A., Trochimowicz, H. J., Terrill, J. B., and Mullin, L. S. (1973). Blood levels of fluorocarbon related to cardiac sensitization. Am. Ind. Hyg. Assoc. J. 34, 102109.[ISI][Medline]
Brown, R. B., Delp, M. D., Lindstedt, S. L., Rhomberg, L. R., and Beliles, R. P. (1997). Physiological parameter values for physiologically based pharmacokinetic models. Toxicol. Ind. Health, 13, 407484.[ISI][Medline]
Chang, K. and Chiou, W.L. (1976). Fluorocarbon aerosol propellants VIII: Solubility of trichloromonofluoromethane in dog blood and tissue homogenates. J. Pharm. Sci. 65, 5355.[ISI][Medline]
Chiou, W. L., and Niazi, S. (1973). A simple and ultra-sensitive head-space gas chromatographic method for the assay of fluorocarbon propellants in blood. Res. Comm. Chem. Pathol. Pharmacol. 6, 481498.[ISI][Medline]
Clarke, K. W., Alibhai, I. K., Lee, Y. H., and Hammond, R. A. (1996). Cardiopulmonary effects of desflurane in the dog during spontaneous and artificial ventilation. Res. Vet. Sci. 61, 8286.[ISI][Medline]
Eger, E. I., II (1990). Uptake and distribution. In Anesthesia, 3rd ed. (R. D. Miller, Ed.), pp. 85104. Churchill Livingstone, New York.
Frei, F. J., Thomson, D. A., and Zbinden, A. M. (1988). Influence of ventilatory and circulatory changes on the pharmacokinetics of halothane and isoflurane. Experientia 44, 178181.[ISI][Medline]
Hellebrekers, L. J. (1986). Comparison of isoflurane and halothane as inhalation anaesthetics in the dog. Vet. Q. 8, 183188.[ISI][Medline]
Holaday, D. A. (1977). Absorption, biotransformation, and storage of halothane. Environ. Health Perspect. 21, 165169.[ISI][Medline]
Holaday, D. A., Fiserova-Bergerova, V., Latto, I. P., and Zumbiel, M. A. (1975). Resistance of isoflurane to biotransformation in man. Anesthesiology 43, 325332.[ISI][Medline]
Hughes, R. L., Campbell, D., and Fitch, W. (1980). Effects of enflurane and halothane on liver blood flow and oxygen consumption in the greyhound. Br. J. Anaesth. 52, 10791086.[Abstract]
Steward, A., Allott, P. R., and Mapleson, W. W. (1975). The solubility of halothane in canine blood and tissues. Br. J. Anaesth. 47, 423434.[Abstract]
Sutton, T. S., Koblin, D. D., Gruenke, L. D., Weiskopf, R. B., Rampil, I. J., Waskell, L., and Eger, E. I., II. (1991). Fluoride metabolites after prolonged exposure of volunteers and patients to desflurane. Anesth. Analg. 73, 180185.[Abstract]
Thomas, V. (1975). BiologicalMathematical Modeling of Chronic Toxicity. AMRL-TR-755. Aerospace Medical Research Laboratory, Wright-Patterson Air Force Base, OH.
U. S. EPA. (1994). SNAP Technical Background Document: Risk Screen on the Use of Substitutes for Class I Ozone-depleting Substances, Fire Suppression and Explosion Protection (Halon Substitutes). U. S. Environmental Protection Agency, Office of Air and Radiation, Stratospheric Protection Division, Washington, DC.
Vinegar, A., and Jepson, G. W. (1996). Cardiac sensitization thresholds of halon replacement chemicals predicted in humans by physiologically based pharmacokinetic modeling. Risk Anal. 16, 571579.[ISI][Medline]
Vinegar, A., Jepson, G. W., Cisneros, M., Rubenstein, R., and Brock, W. J. (2000). Setting safe acute exposure limits for halon replacement chemicals using physiologically based pharmacokinetic modeling. Inhal. Toxicol. 12, 751763.[ISI][Medline]
Vinegar, A., Jepson, G. W., and Overton, J. H. (1998). PBPK modeling of short-term (0 to 5 min) human inhalation exposures to halogenated hydrocarbons. Inhal. Toxicol. 10, 411429.[ISI]
Vinegar, A., Winsett, D. W., Andersen, M. E., and Conolly, R. B. (1990). Use of a physiologically based pharmacokinetic model and computer simulation for retrospective assessment of exposure to volatile toxicants. Inhal. Toxicol. 2, 119128.
Williams, R. J., Vinegar, A., McDougal, J. N., Jarabek, A. M., and Fisher, J. W. (1996). Rat to human extrapolation of HCFC-123 kinetics deduced from halothane kinetics: A corollary approach to physiologically based pharmacokinetic modeling. Fundam. Appl. Toxicol. 30, 5566.[ISI][Medline]
|