Derivation of a Human Equivalent Concentration for n-Butanol Using A Physiologically Based Pharmacokinetic Model for n-Butyl Acetate and Metabolites n-Butanol and n-Butyric Acid

J. G. Teeguarden*,1, P. J. Deisinger{dagger}, T. S. Poet*, J. C. English{dagger}, W. D. Faber{ddagger}, H. A. Barton§, R. A. Corley* and H. J. Clewell, III

* Battelle, Pacific Northwest National Laboratory, Richland, Washington 99352; {dagger} Health and Environmental Laboratories, Eastman Kodak Company, Rochester, New York 14652-6272; {ddagger} Willem Faber Toxicology Consulting, Victor, New York 14564; § NHEERL, USEPA, Research Triangle Park, North Carolina 27709; and ENVIRON Health Sciences Institute, Ruston, Louisiana 71270

1 To whom correspondence should be addressed at Pacific Northwest National Laboratory, PO Box 999, Mail Stop P7-56, Richland, WA 99352. Fax: (509) 376-9449. E-mail: justin.teeguarden{at}pnl.gov.

Received September 7, 2004; accepted December 20, 2004


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
The metabolic series approach for risk assessment uses a dosimetry-based analysis to develop toxicity information for a group of metabolically linked compounds using pharmacokinetic (PK) data for each compound and toxicity data for the parent compound. The metabolic series approach for n-butyl acetate and its subsequent metabolites, n-butanol and n-butyric acid (the butyl series), was first demonstrated using a provisional physiologically based pharmacokinetic (PBPK) model for the butyl series. The objective of this work was to complete development of the PBPK model for the butyl series. Rats were administered test compounds by iv bolus dose, iv infusion, or by inhalation in a recirculating closed chamber. Hepatic, vascular, and extravascular metabolic constants for metabolism were estimated by fitting the model to the blood time course data from these experiments. The respiratory bioavailability of n-butyl acetate (100% of alveolar ventilation) and n-butanol (50% of alveolar ventilation) was estimated from closed chamber inhalation studies and measured ventilation rates. The resulting butyl series PBPK model successfully reproduces the blood time course of these compounds following iv administration and inhalation exposure to n-butyl acetate and n-butanol in rats and arterial blood n-butanol kinetics following inhalation exposure to n-butanol in humans. These validated inhalation route models can be used to support species and dose-route extrapolations required for risk assessment of butyl series family of compounds. Human equivalent concentrations of 169 ppm and 1066 ppm n-butanol corresponding to the rat n-butyl acetate NOAELs of 500 and 3000 ppm were derived using the models.

Key Words: PBPK model; pharmacokinetics; acetates; extrapolation; risk assessment; metabolic series approach; n-butyl acetate; n-butanol; n-butyric acid.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
n-Butyl acetate, n-butanol, n-butyraldehyde, and n-butyric acid (the butyl series) are a family of metabolically related2 compounds also linked by an industrial production method relying on sequential oxidation/reduction reactions. In order to make toxicity testing and the development of acceptable exposure concentrations as efficient as possible, a methodology which takes advantage of the metabolic relationship between the butyl series compounds, called the metabolic series approach (previously, "the family approach"), was developed (Barton et al., 2000Go). The metabolic series approach uses a dosimetry-based analysis to develop toxicity information for a group of metabolically linked compounds using pharmacokinetic data for each compound and toxicity data for the parent compound (Barton et al., 2000Go). In the case of the butyl series, exposure to the parent ester (n-butyl acetate) leads to systemic exposure to each of the three sequential metabolites, which are produced in stoichiometric yield. Exposure to n-butanol leads to systemic exposure to each of the two subsequent metabolites. Results of exposures with a single compound can be extrapolated to subsequent metabolites and from one dose route to another (e.g., oral to inhalation). Dose-route and interspecies extrapolations (e.g., rat inhalation to human inhalation) can be made with a physiologically based pharmacokinetic (PBPK) model.

For such an approach to be effective, the PBPK model must be developed, and where possible, verified for exposure routes necessary for interpretation of toxicity studies in test animals and for routes for which there is the expectation of significant human exposure. This includes accurate characterization of absorption, distribution, and metabolic and renal clearance.

For both n-butyl acetate and n-butanol, the primary exposure route of concern for humans is inhalation; exposure to n-butyric acid is of lower concern, in part because the low odor threshold for this compound may be exposure limiting. Rat inhalation toxicity studies are available for butyl acetate (David et al., 1998Go, 2001Go). These studies result in co-exposure to n-butanol in rats, raising the possibility of identifying rat n-butanol inhalation exposures that would result in equivalent blood n-butanol concentrations. Consequently, there was a need identified to parameterize and validate a rat model for the inhalation exposure to both n-butyl acetate and n-butanol. The target tissue dose-effect relationships from the rodent toxicity studies could be extrapolated to humans with the use of the PBPK model parameterized for the human.

At the time of publication of the provisional butyl series model (Barton et al., 2000Go), the limitations of the existing pharmacokinetic (PK) data had been identified, and the initial model was available for assisting in the development and evaluation of additional experimental data required for parameterization and, in some cases, verification of the model. This included PK data to refine the characterization of systemic clearance of the butyl series compounds, and respiratory absorption (and bioavailability) following inhalation of n-butyl acetate and n-butanol in the rat. Clearance in the human would be based on principles of allometric scaling, and literature values for absorption following inhalation would be used for n-butyl acetate and n-butanol.

The objective of this work was to complete the development of rat and human PBPK models for butyl series compounds based on the intravenous (iv) kinetics of n-butyl acetate, n-butanol and n-butyric acid in the rat, and inhalation kinetics of n-butyl acetate (rat only) and n-butanol (rats and humans). In addition the PBPK model was used to estimate inhalation-route n-butanol human equivalent concentrations based on rat inhalation route NOAELs.


    MATERIAL AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
Experimental Data
Data availability.
Data from six unpublished reports are utilized for model development (Corley et al., 2000Go; Deisinger and English, 2000Go, 2001Go, 2003Go; Poet et al., 2003aGo,bGo). All studies involving animals were IACUC approved. Experimental details not provided here may be found in the reports, which are available by request from the sponsor (American Chemistry Council, 1300 Wilson Blvd., Arlington, VA 22209).

Rat iv bolus and iv infusion blood kinetics studies (Deisinger and English, 2001).
Young-adult, male Sprague-Dawley rats (280–330 g, Zivic-Miller Laboratories, Zelienople, PA) were surgically prepared by the vendor with femoral and jugular vein cannulae (Deisinger and English, 2001Go). Pilot studies were conducted to identify acutely nontoxic iv doses for n-butyl acetate, n-butanol, and n-butyric acid, verify the efficacy of the analytical methods, determine background concentrations for each of the analytes, and establish sampling times for the definitive studies. Pilot and definitive iv bolus studies were conducted using the same protocol. Groups of rats (five/group for definitive studies, one/group for pilot studies) were administered a bolus injection of each test chemical dissolved in a saline carrier (0.9% saline; plus 1% Tween 20 vehicle for n-butyl acetate) at ~0.28 and ~0.028 (n-butyric acid only) mmol/kg via the femoral vein cannula. Exact administered doses are presented in figure legends. In addition, n-butyric acid was delivered by iv infusion, ~0.28 mmol/kg delivered over 3 min, using a syringe pump (Kd Scientific Model 100, Boston, MA). Sample collection and processing was the same for bolus and infusion experiments. Aliquots of blood (100 µl) were obtained by serial sampling from the jugular vein cannula. The blood was immediately spiked with known amounts of hexanoic acid and isoamyl alcohol as internal standards, de-proteinized with 0.4 M H2SO4, extracted with diethyl ether, methylated, and analyzed by GC-MS. Mean data are presented as points in figures with error bars corresponding to two times the SE, equivalent to the 95% confidence limits on the mean value. Background concentrations, collected pre-dosing, were established for each compound. Results from the definitive studies were used for model development while those of the pilot study data were used to verify the model parameterization.

Rat Closed Chamber Exposure and Blood Kinetics (Poet et al., 2003a)
Animals and chamber.
Male Sprague-Dawley rats with indwelling jugular cannulae were purchased from two vendors. Rats used in the n-butanol studies (average body weight 317 g) and one group of two animals (5 and 8, average body weight 305 g) used in the n-butyl acetate studies were obtained from Hilltop Lab (Scottsdale, PA). A second group of four animals (average body weight 354 g) was obtained from Charles River (Raleigh, NC). These animals were significantly less active that the Hilltop Lab animals. The animals were housed in suspended plastic cages with chipped bedding and acclimated to the laboratory for at least three days prior to exposure. Gas uptake methods for several volatile water-soluble compounds, including n-butyl acetate, were developed as described previously (Corley et al., 2000Go). These methods were used in conjunction with plethysmography to measure both the gas uptake and respiratory parameters of the rats during exposures to target concentrations of ~2000 ppm n-butyl acetate (Poet et al., 2003aGo) or n-butanol (Poet et al., 2003bGo). n-Butyl acetate was obtained from Sigma (St. Louis, MO, lot no. 13091CN ) and was 99.7% pure; n-butanol (lot no. 952729), obtained from FisherChem (Fairlawn, NJ), was 99.8% pure.

The gas uptake chamber used in previous experiments (Gargas et al., 1986Go) was modified to allow the introduction of the plethysmograph and its associated connections to a computer monitoring system, and to accommodate a cannula that was accessible for blood draws during the exposure (Supplemental Data, Fig. S1). The standard glass lid was replaced with a stainless steel lid. The volume of the closed chamber and restraint tube was 9.44 l. The restraint tube was included in the empty chamber simulations to capture losses associated with its presence. Oxygen was maintained at 20 ± 1%. Previous studies in this laboratory indicated that significant losses of test materials occurred with the use of a CO2 scrubber and water condensation from an ice-bath normally used to condense water vapor generated by the animals (Corley et al., 2000Go). Accordingly, the scrubber was removed, no ice bath was included, and chamber runs were limited to a single rat for 2 h.

Exposure and plethysmography.
A whole-body plethysmograph (designed and built at Battelle, Toxicology Northwest) linked to a Buxco Biosystem XA Data Acquisition System (Buxco Electronics, Inc. Sharon, CT) was used for non-invasive measurements of tidal volume (ml), respiratory rate and minute ventilation (ml/min) on conscious rats. The animals were restrained in a constant-pressure plethysmograph with an attached pneumotachograph. A neck seal separated the head chamber from the body chamber and ventilation parameters were computed from flow measurements through the pneumotachograph from the body chamber.

The rats were acclimated to a tube that was similar to the plethysmograph for 30 min/day for three days prior to the exposure. The actual restraint plethysmograph tube was not used for acclimation due to difficulties with the cannulation surgery site and the neck restraint system. Exposures included: (1) empty chamber to determine the loss to the system, (2) an individually exposed dead rat to determine the adsorptive losses of test material to the head of the rat, and (3) individually exposed live rats to determine the respiratory clearance from the gas uptake chamber while concurrently collecting respiratory parameters and blood samples for analysis of compound (n-butyl acetate or n-butanol) and its metabolites. Each rat was weighed and exposed to target concentrations of 2000 ppm of n-butyl acetate (Poet et al., 2003aGo) or n-butanol (Poet et al., 2003bGo) for 2 h in the recirculating inhalation chamber. Initial chamber concentrations varied from ~1400–2000 ppm.

Chamber monitoring.
The concentration of n-butyl acetate or n-butanol was monitored using a Hewlett Packard 5890 Series II capillary gas chromatograph with flame ionization detection (GC/FID; Hewlett Packard, Palo Alto, CA). A 30 m x 0.32 mm ID DB-624 capilary column with 3 µm film thickness was used for separation (J&W Scientific, Folsom, CA). The concentration was monitored approximately every 7 min via automatic injections of ~1 ml of chamber air via a Valco gas sampling valve. The GC/FID was calibrated with a standard curve consisting of standards (n-butanol or n-butyl acetate) prepared in 3-L Tedlar bags (SKC, Fullerton, CA) prior to each exposure.

Respiratory measurements.
The respiratory rates were determined from the peak of the inspiratory portion of the breathing cycle. The respiratory flow was sampled at a minimum frequency of 400 Hz. All calculated respiratory parameters were averaged for each 1-min portion of the measurement period. Data were collected for 10 min prior to the start of the exposure period to be used as the baseline for each animal. Butyl acetate exposed animals were divided into two groups for purposes of simulation: a four-animal group (Charles River, animals 1–4) with body weights higher than those of animals used for parameterizing the model, and a second group (Hilltop Lab, animals 5–8) with body weights in the range of those used for model parameterization. Another distinction between these groups was that the four-animal group had lower ventilation rates and was exposed to higher initial chamber concentrations (1625 ppm vs. 1400 ppm) than the two-animal group (averaged post-experiment).

Blood analysis methods.
Blood was placed in pre-weighed 4-ml screw cap vials containing 2 µg of phenyl methyl sulfonyl fluoride (PMSF) to inhibit any potential further metabolism of n-butylacetate. The vials were quickly sealed, weighed, and placed on dry ice. The blood was stored at –80°C until analysis. A weighed quantity (~0.10 g) of blood was extracted with 200 µl of 0.9 M H2SO4, 0.25 g Na2SO4 (to improve extraction efficiencies), and 200 µl ethyl acetate containing known concentrations of internal (matrix) standards (hexyl alcohol and ethoxyacetic acid). Separation and quantitation were achieved using a 60 m x 0.32 mm id x 1.0 film thickness Stabilwax DA capillary column (Restek, Bellefonte, PA) and a Hewlett Packard 6890 GC with flame ionization detection (FID). Splitless injections of 2.0 µl of extract at an injector temperature of 210°C were employed. The initial oven temperature was 60°C for 1.5 min then increased at 15°C/min to 250°C. Hydrogen was used as the carrier gas at 25 psi for 11 min then increased at 50 psi/min to 35 psi. The flame ionization detector (FID) temperature was 275°C. For the preceding conditions, n-butyl acetate, n-butanol, hexyl alcohol, n-butyric acid, and ethoxyacetic acid had retention times of 5.4, 6.1, 8.6, 11.3, and 13.3 min, respectively.

Rat open chamber n-butyl acetate exposure and blood kinetics (Groth and Freundt, 1991).
The intratracheal inhalation experiment of Groth and Freundt (1991)Go provides estimates of the uptake of n-butyl acetate by the lungs, its metabolism to n-butyl alcohol, and the rate of clearance of n-butyl alcohol. This published study used anesthetized female Sprague-Dawley rats (290–340 g) exposed to air containing an average concentration of 970 ppm n-butyl acetate for 5 h. Arterial blood samples were taken periodically and analyzed for n-butyl acetate and n-butyl alcohol; data were digitized from graphs in the original manuscript, as reported previously (Barton et al., 2000Go). Since the exposure was by tracheostomy tube, no presystemic nasal metabolic clearance would occur; the concentration presented to the lungs was equal to the measured air concentration. These air concentrations were included in a table function within the modeling software (Advanced Continuous Simulation Language, Aegis, Inc., Huntsville, AL), and interpolated values were used as the instantaneous inhalation concentrations for simulating the 5-h exposure.

Inhalation route blood kinetics of n-butanol in humans (Astrand et al., 1976).
Two groups of six male subjects, average body weight of 78.3 kg and ages between 21 and 34 years, were exposed to n-butanol through a breathing valve and mouthpiece (Astrand et al., 1976Go). n-Butanol exposure concentrations were either 100 or 200 ppm. Exposures were conducted during rest and at various levels of exercise. Blood samples were taken through catheters placed into a brachial artery and a medial cubital vein (both in the arm). The bioavailability of n-butanol, as a fraction of pulmonary (minute ventilation) ventilation, was estimated by the authors as the difference between the amount given (concentration x ventilation rate) and the amount in expiratory air (total amount collected in the Douglas bag). Defined this way, bioavailability is the fraction of material in the inhaled volume of air reaching the blood exchange region in the lung. Respiratory bioavailability can be reported as fraction of either pulmonary (minute ventilation) or alveolar ventilation. Pulmonary ventilation rates were estimated from the volume of expired air collected in the Douglas bag.

The average bioavailability was 40% of pulmonary ventilation or 59% of alveolar ventilation assuming all absorption into systemic blood occurs in the alveolar region; alveolar ventilation rates are used in the model. Mean pulmonary ventilation rates, blood concentrations and fractional uptakes (inhaled-exhaled n-butanol over a 30 minute period) were reported for each exposure condition (see Supplemental Data, Table S1). Pulmonary ventilation rates were converted to alveolar ventilation rates for use in the PBPK model by assuming alveolar ventilation rates are ~67% of pulmonary ventilation rates (Brown et al. 1997Go). Arterial blood concentrations in a single individual were also reported for a 90-min period following the end of exposure to 100 ppm n-butanol. Arterial and venous blood concentrations of n-butanol did not fall below 1.1 µM after the end of exposure, suggesting this is near background concentrations. This background concentration was subtracted from the reported blood concentrations; only exogenous n-butanol was simulated by the model.

Model structure.
The model structure and governing equations were as reported by Barton et al. (2000Go) with some revisions. The model was coded using ACSL (Advanced Continuous Simulation Language, Aegis, Inc., Huntsville, AL) and is available upon request. Briefly, the model is composed of four linked submodels, one each for n-butyl acetate, n-butanol, n-butyraldehyde, and n-butyric acid (Fig. 1). Each submodel is constructed of four or five compartments required for absorption, storage, metabolism, distribution, and clearance. The compartments include lung, arterial and venous blood, fat (n-butyl acetate only), liver, and the remaining perfused tissues (other perfused tissues, OPT) (Fig. 1). All compartments are described as well mixed and flow limited. The upper respiratory tract compartments described in Barton et al. (2000)Go—nose, trachea, conducting airways—were removed for the simulations reported here. To account for fractional bioavailability of water soluble volatiles such as n-butanol, the incoming air concentration (conc) was multiplied by a factor (fa) representing the fraction of chemical that would be available for absorption (Barton et al., 2000Go). The model was further revised to run in both closed and open chamber inhalation conditions used in the experiments reported by Poet et al. (2003aGo,bGo), and a provision was included to run simulations based on reported study-specific ventilation rates, coded in an ACSL table function. Michaelis-Menten metabolism of n-butyl acetate and n-butyric acid, and first order metabolism of n-butanol was added to the other perfused tissue compartment (Fig. 1), reflecting significant extrahepatic, extravascular metabolism implied by earlier simulations. Simple first order metabolism was assumed for n-butanol in the other perfused tissue compartment in the absence of sufficient data to implement a Michaelis-Menten approach. First order and Michaelis-Menten metabolism are formulated as described in Barton et al. (2000Go).



View larger version (22K):
[in this window]
[in a new window]
 
FIG. 1. The butyl series model is composed of four linked submodels, one each for n-butyl acetate, n-butanol, n-butyraldehyde, and n-butyric acid. Each submodel is constructed of four or five compartments, required for absorption, storage, metabolism, distribution, and clearance and include compartments describing the lung, combined arterial and venous blood, fat (n-butyl acetate only), liver and the other perfused tissues.

 
Development of the rat PBPK model.
The various model parameters were either obtained from the literature, or inferred by visual fitting to in vivo PK data. Values and sources for tissue volumes, blood flows, ventilation rates, partition coefficients, and some metabolic constants were reported previously (Barton et al., 2000Go). The sources of new parameters are described here, with a complete list of parameters used in these simulations provided in Tables 1–3GoGo.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Model Physiological Parameters

 

View this table:
[in this window]
[in a new window]
 
TABLE 2 Metabolic Constants for Butyl Series Compounds

 

View this table:
[in this window]
[in a new window]
 
TABLE 3 Partition Coefficients Calculated from Kaneko et al. (1994)Go

 
Intravenous blood time course data for each compound and its subsequent metabolite(s) (Deisinger and English, 2001Go) were used to estimate metabolic constants (Vmax and Km or Kmet) for B-carboxylesterase (CE), alcohol dehydrodgenase (ADH), aldehyde dehydrogenase (ALDH), and medium chain acyl-CoA synthetase (mACS), the latter capable of, and presumed responsible for, clearance of n-butyric acid (Knights, 1998Go). Some general tissue distribution (e.g., the need for activity in tissues other than the liver) information for CE and ADH was also inferred from the blood data. The Km for CE metabolism of n-butyl acetate was assumed to be equivalent to that of ethyl acetate (Deisinger and English, 2003Go): for liver the Km was set to 1 nM, and for blood and other perfused tissues the value was set to 100 nM. This revision of the Km values initially estimated from fitting a more limited PK data set (Barton et al., 2000Go) was required to fit peak concentrations of n-butyl acetate measured in the more refined PK studies presented here. Other metabolic constants were fit (as indicated in Table 2) using blood time course data for each compound (Deisinger and English, 2001Go). Model simulations of blood concentrations were visually fit to the observed data. Butyraldehyde is not observed in blood; metabolic constants for butyraldehyde were fixed to simulate the appearance of n-butyric acid following n-butanol or n-butyl acetate dosing. Differential sensitivities of the distribution- and clearance-dominated phases of the blood concentration time-course to the various model parameters allowed determination of vascular, and extra-vascular metabolic constants for each compound. The definitive iv study was used to parameterize the model, and the pilot studies were used to test the model. The respiratory bioavailability study was not used to parameterize the model, rather, given fitted parameters estimated from the iv data, these data were used solely to estimate bioavailability.

The rate and extent of urinary elimination of n-butanol is unknown. Studies with 2-butanol found urinary clearance of up to 14% in different species, so a value of 30 (h)–1 was used for the n-butanol urinary clearance rate (kfilt2) resulting in about 10% urinary excretion (Dietz et al., 1981Go). Urinary excretion data were not available for n-butyric acid, so it was assumed to be less than 3.5%, as observed for isobutyric acid in the rat (DiVincenzo and Hamilton, 1979Go). Rapid metabolic clearance of n-butyl acetate is expected to limit urinary elimination. The value estimated for n-butyric acid was also used for n-butyl acetate.

Development of the human PBPK model.
Physiological parameters such as flow rates and tissue volumes were obtained from the literature and are reported in Table 1. The fractional availability of n-butanol (59% of alveolar ventilation) was used as reported by Astrand et al. (1976)Go. Rate constants for metabolism and renal elimination were allometrically scaled from the rat values, and are reported in Tables 1 and 2. In the absence of better data, the rat partition coefficients were assumed to be a good proxy for human values, and used directly in the human model. The human model was validated only for the inhalation route for n-butanol.

Modeling approach.
Model development for the rat proceeded in two phases and concluded with verification of predictions of blood concentrations of n-butyl acetate, n-butanol, and n-butyric acid following inhalation exposures to n-butyl acetate and n-butanol as reported by Poet et al. (2003aGo,bGo). First, the model structure was evaluated and kinetic constants for metabolic clearance (and production) of each metabolic series member was estimated by fitting the model to the corresponding blood kinetics for each compound following iv bolus or iv infusion. The parameterization was verified through simulation of additional iv blood kinetics data. Following the establishment of a convincing parameterization for the systemic fate of the butyl series compounds (based on fitting the model to the iv blood kinetics data), the inhalation route was parameterized. Parameterization was achieved by estimating n-butyl acetate and n-butanol bioavailability, the final parameter, through simulation of closed chamber data. Validation of the inhalation route was conducted by comparison of simulated and observed blood concentrations of butyl series compounds following inhalation of n-butyl acetate and n-butanol. The results of these analyses are presented in the order: parameterization of systemic metabolism using iv studies, verification of systemic metabolism, estimation of bioavailability, and verification of model prediction of blood concentrations of butyl series compounds following inhalation exposure. A list of the experimental studies and their use in developing and validating the model are provided as supplemental data (see Supplemental Data, Table S2).

The human model was evaluated by simulating the blood n-butanol kinetics following n-butanol inhalation (Astrand et al., 1976Go) without changes in parameters. The human model was validated only for the inhalation route for n-butanol.

Sensitivity and uncertainty analysis.
The impact of uncertainty in the various model parameters on simulated arterial blood concentrations was evaluated semi-quantitatively though a combination of sensitivity analysis and a qualitative evaluation of parameter uncertainty. Each parameter was characterized as having low, medium, or high sensitivity and low, medium, or high uncertainty. The sensitivity of the arterial n-butanol concentrations to all model parameters was evaluated and those with sensitivity coefficients greater than 0.1 were also reported. The sensitivity coefficient (SC) is defined as the percent change in the dose metric for a 1% change in the listed parameter. A sensitivity coefficient of 1 indicates that there is a 1 to 1 relationship between the change in the parameter and the internal dose metric. Negative sensitivity coefficients indicate an inverse relationship between the model parameter and dose metric. Based on the sensitivity analysis, each parameter was characterized as having a low, medium or high impact on the selected dose metric using the following criteria:

Low (L): SC between 0.1 and 0.15
Medium (M): SC between 0.15 and 0.5
High (H): SC greater than or equal to 0.5

Parameters with SC less than 0.1 were not reported or considered further in this analysis.

The uncertainty in parameters with SC greater than 0.1 was evaluated qualitatively. The source of the parameter was reviewed and the parameters were characterized as having low, medium, or high uncertainty based on the following criteria:

Low: Obtained from data in the correct species, coefficient of variation (CV) less than 0.5 or verified through successful use in published PBPK models
Medium: Correct species, CV greater than 0.5, or scaled value from a different species with a high probability that scaling holds across species
High: All others

The sensitivity and uncertainty designations for each parameter were tabulated and used to identify parameters with sufficient uncertainty (uncertainty designation greater than Low) and impact on arterial n-butanol concentrations (sensitivity designation greater than Low) to have some influence on the dose-route extrapolation and risk assessment.

Sensitivity analyses for the rat and human PBPK models, measuring the impact of model parameters on blood n-butanol concentrations, were run for n-butyl acetate (rat only) and n-butanol concentrations appropriate for testing the model in the range it will be exercised for dose-route extrapolations supporting risk assessment. Sensitivity analyses were conducted for two steady state n-butyl acetate concentrations, 500 ppm and 3000 ppm, the NOAELs (not duration adjusted) for reduced weight gain and neurotoxicity, respectively, in rat 90-day subchronic studies (David et al., 1998Go, 2001Go). Sensitivity analyses for the rat and human for constant n-butanol exposures were conducted at 700 and 1000 ppm, respectively. These n-butanol concentrations approximate exposures resulting in blood n-butanol concentrations near to those observed for the rat at the NOAEL of 500 ppm n-butyl acetate.

Human equivalent concentrations.
The human n-butanol exposure concentration resulting in a weekly arterial blood n-butanol AUC (µM x h) equivalent to the weekly blood n-butanol AUC (µM x h) in the rat following NOAEL concentration exposure (6 h/day, 5 days/week) is termed the human equivalent concentration (HEC). The HEC was determined for two rat NOAELs, 500 ppm and 3000 ppm, the NOAELs for reduced weight gain and neurotoxicity, respectively, in rat 90-day subchronic studies (David et al., 1998Go, 2001Go). The HEC was determined by first using the rat model to determine the a weekly arterial blood n-butanol AUC associated with NOAEL concentration exposures and then using the human model to determine the human n-butanol exposure concentration leading to the same weekly AUC under continuous exposure conditions. The rat and human ventilation rates used in these simulations are found in Table 1.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
iv Kinetics of the Butyl Series Compounds in Rats
Experimental results (Deisinger and English, 2001).
The background concentrations of the butyl series analytes in rat blood were determined over the course of a day at hourly intervals from three non-dosed, dual cannulated, male SD rats. Blood samples were collected from the jugular vein cannula and were analyzed by GC-MS. Background n-butyric acid concentrations were near or below the limit of detection (0.61 µM), and we chose to represent background levels in figures with a value of 1 µM. Trace amounts of n-butanol (less than 3 µM), below the limit of quantitation (2.15 µM), were detected in two of 21 rat blood samples assayed. N-butyl acetate (limit of detection, 1.43 µM) and n-butyraldehyde (limit of detection, 7.0 µM) were not detected in these samples.

In vivo hydrolysis of the n-butyl acetate ester was very rapid, with a half-life measured in seconds, rather than minutes or hours. n-Butanol, a product of this hydrolysis, has a somewhat slower elimination rate, but is present only in trace amounts in the blood by 30 min post dosing with either n-butyl acetate or n-butanol. The terminal butyl series oxidation product, n-butyric acid, appears to have a half-life of less than 1 min when it is administered in vivo, but its elimination following administration of n-butyl acetate or n-butanol is extended by the concurrent production of the compound from its precursors.

Simulation of iv route blood kinetics.
Parameterization of metabolic constants for downstream metabolites of n-butyl acetate is complicated by blood concentration time course dependence on the metabolism of the upstream metabolite, which controls the production rate. Therefore, metabolic constants were fitted first for n-butyric acid, followed by n-butanol and n-butyl acetate. Blood concentration time course data for metabolites (n-butanol, n-butyric acid) following administration of a parent compound (n-butyl acetate, n-butanol) were also utilized for parameter estimation. All data are from Deisinger and English (2001)Go.

Following iv bolus and iv infusion administration of n-butyric acid, modeled blood concentrations of n-butyric acid were in close agreement with observed values, falling within or just outside the 95% confidence limits (Fig. 2). Because the model does not consider endogenous sources of n-butyric acid, the simulation line drops below observed concentrations after blood concentrations decline to background concentrations at 9 min (Fig. 2). The 10 and 15 min data points in the bolus study were close to the background concentrations and assumed not to reflect a terminal elimination phase different than the suggested by the iv infusion data.



View larger version (13K):
[in this window]
[in a new window]
 
FIG. 2. Simulated (solid lines) and mean observed (symbols) blood n-butyric acid concentrations following iv bolus (0.279 mmoles/kg, {blacksquare}, five animals, average body weight of 0.3 kg) or i.v. infusion (0.281 mmoles/kg, •, one animal, body weight of 0.33 kg) of n-butyric acid. Close agreement between simulated and observed data implies proper characterization of distribution and metabolic clearance of n-butyric acid. Horizontal line (.......) represents background concentrations. Error bars represent ± 2 SE, equivalent to the 95% confidence limits on the mean value.

 
Blood kinetics following iv administration of n-butanol were also well described by the model, with simulations falling within the 95% confidence limits for most of the time course. Peak concentrations, as well as the distribution- and clearance-dominated phases of the n-butanol blood concentration time course, were in close agreement with observed values (Fig. 3A). Blood concentrations of n-butyric acid were also well described by the model, but peak concentration was modestly under-predicted (Fig. 3B).



View larger version (16K):
[in this window]
[in a new window]
 
FIG. 3. n-Butanol iv bolus administration (0.279 mmoles/kg): Simulated (solid lines) and mean observed (symbols, five animals, average body weight, 0.28 kg) blood concentrations of (A) n-butanol and (B) n-butyric acid. Error bars represent ± 2 SE.

 
The time (Tmax) and magnitude (Cmax) of the peak blood concentration for n-butyl acetate were sensitive to blood Vmax/Km and the iv delivery time (2–3 s), while the initial rapid clearance phase was sensitive to blood Vmax/Km, and the slower terminal elimination phase was most sensitive to extra-vascular metabolism. Metabolic clearance of n-butyl acetate is rapid following iv administration (Fig. 4). Sampling times were long relative to Tmax, but modeled and observed blood concentrations for the distribution/elimination phase were in reasonable agreement (Fig. 4A). As with n-butyric acid, significant extrahepatic metabolism of the n-butyl acetate was indicated, and in addition to vascular metabolism, it was necessary to revise the model, attributing extravascular metabolism to the other perfused tissue compartment. Modeled and observed concentrations of n-butanol and n-butyric acid were consistent with observed values following iv administration of n-butyl acetate (Figs. 4B, 4C), although peak concentrations of n-butyric acid were under-predicted by the model.



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 4. n-Butyl acetate iv bolus administration (0.282 mmoles/kg): Simulated (solid lines) and mean observed (symbols, five animals, average body weight, 0.30 kg) blood concentrations of (A) n-butyl acetate, (B) n-butanol, and (C) n-butyric acid. Error bars represent ± 2 SE.

 
Validation of systemic metabolism.
Parameterization of the model was verified by simulation of the blood kinetics of n-butyl acetate, n-butanol, and n-butyric acid following iv administration of n-butyl acetate to a single animal in a pilot study (Deisinger and English, 2001Go). The agreement between modeled and observed blood concentrations of these compounds provides confirmation that their systemic distribution and metabolic clearance are adequately described by the model (Fig. 5).



View larger version (18K):
[in this window]
[in a new window]
 
FIG. 5. n-Butyl acetate iv bolus administration (0.282 mmoles/kg, body weight 0.33 kg): Simulated (solid lines) and observed (symbols) blood concentrations of (A) n-butyl acetate, (B) n-butanol, and (C) n-butyric acid. Results from a single animal used to verify model parameterization.

 
Inhalation Kinetics of the Butyl Series Compounds in Rats
Experimental data.
Average measured minute ventilation rates during inhalation exposure to ~2000 ppm n-butyl acetate and n-butanol (Poet et al., 2003aGo,bGo) are shown in Figures 6A and 6B. Ventilation rates for animals or groups of animals in the simulations presented here are provided as supplemental data (see Supplemental Data, Ventilation Rates). During both n-butyl acetate and n-butanol exposure, there was an initial respiratory depression followed by an increase in minute ventilation rates over the length of the exposure period. Loss of n-butyl acetate in the empty apparatus with (Fig. 7A, dead rat) and without (data not shown) a nonrespiring rat, was the same. Loss attributable to the chamber apparatus was large relative to that attributable to respiratory uptake (Fig. 7A). Nonetheless, respiratory uptake of n-butyl acetate was both quantifiable and sufficient for measuring respiratory bioavailability. This was also true for n-butanol (Fig. 7B, dead rat data not shown).



View larger version (38K):
[in this window]
[in a new window]
 
FIG. 6. Average (± SD) minute volume from six animals exposed via inhalation to n-butyl acetate (Panel A; Poet et al., 2003aGo) or to n-butanol (Panel B; Poet et al., 2003bGo).

 


View larger version (22K):
[in this window]
[in a new window]
 
FIG. 7. Panel A: Simulated (lines) and observed (data points) chamber loss of n-butyl acetate (dead rat only, {blacksquare}) and respiratory uptake (•) of n-butyl acetate. Chamber loss with and without a nonrespiring rat was the same. The respiratory uptake simulation reflects the average of four independent experiments; average chamber concentrations, average body weights and average alveolar ventilation rates were used. Uptake was well described when the respiratory bioavailability of butyl acetate was assumed to be 100% of alveolar ventilation. Panel B: Respiratory uptake of n-butanol is shown for four separate rat experiments for which different values of respiratory bioavailability were required to fit the data. The bioavailability values used were 40, 60, 60, and 40% for rats 3, 5, 7, and 8. Body weights corresponding to these animals were, 0.30, 0.30, 0.33, and 0.32 kg, respectively. The average respiratory bioavailability was 50% of alveolar ventilation for the group. Empty chamber loss is also shown. Lines are model simulations and points represent measured chamber concentrations.

 
The n-butyl acetate peak concentration in the blood appeared to occur near 20 min, though we would expect peak concentrations to occur earlier for this rapidly metabolized compound, and before chamber concentrations had fallen significantly. Blood n-butanol concentrations peaked 20 min after the start of exposure (Figs. 11 and 12). Butyric acid was not detected in any sample. Blood n-butyl acetate and n-butanol concentrations were greater than the lowest concentration in the standard curve, which was the effective limit of reliable quantitation. These values were 3.7 µM and 9.4 µM for n-butyl acetate and n-butanol, respectively. The detection limit for n-butyric acid was 12.5 µM in the n-butyl acetate experiments. During the n-butanol exposures, peak concentrations of n-butanol (~0.1 µM) were observed approximately 20 min after exposure (Fig. 9). N-butyric acid concentrations were near the limit of quantitation (two times the limit of detection, which was lower in the n-butanol experiments than in n-butyl acetate experiments), reaching peak concentrations near 0.007–0.009 µM after 30–45 min.



View larger version (17K):
[in this window]
[in a new window]
 
FIG. 11. Simulated and observed arterial blood concentrations of n-butanol during inhalation of 100 ppm (A, Series II from Astrand et al., 1976Go) or 200 ppm (B, Series I from Astrand et al., 1976Go) n-butanol with various levels of exercise For 100 ppm exposures, adult male volunteers were exposed for 30 min at rest, followed by a 20 min non-exposure period (15 min at rest, 5 min at an exercise rate of 50 watts), followed by three 30 min exposure periods at 50, 100, and 150 watt exercise levels. Post exposure n-butanol concentrations (last four data points) from a single volunteer are also shown. For 200 ppm exposures, male volunteers were exposed for 30 min at rest, followed by a 20 min non-exposure period (15 min at rest, the latter five min at an exercise rate of 50 watts), followed by three 30 min exposure periods at a 50 watt exercise level. Error bars represent ± 2 SE.

 


View larger version (30K):
[in this window]
[in a new window]
 
FIG. 12. Model-derived relationship (points) between rat n-butyl acetate (A) and rat and human n-butanol (B) exposures and arterial blood n-butanol AUCs. The rat AUCs reflect the six h/day, five days/week exposure protocol of the toxicity study and the human AUCs are calculated for continuous exposures, thus accounting for the required exposure duration adjustment. ({diamondsuit}), Rat n-butanol AUC; (x), Human n-butanol AUC. The equations of lines describing the exposure to arterial n-butanol AUC relationship are shown.

 


View larger version (18K):
[in this window]
[in a new window]
 
FIG. 9. n-Butanol closed chamber study: model predicted and observed blood concentrations of n-butanol (A) and n-butyric acid (B). Average concentrations for a group of four animals (5, 6, 7, and 8, average body weight 0.32 kg) are shown. Error bars represent ±2 SE.

 
Inhalation-route bioavailability of n-butyl acetate.
The rate of n-butyl acetate chamber only loss was biphasic (Fig. 7A). Two first order rate constants were fitted to the chamber only (or chamber plus dead rat) data, one describing loss during the first phase and one describing loss during the second phase. The switch between these two rate constants was initiated at the time the rates appeared to change (Fig. 7A). The fitted rate constants described both chamber-only loss and chamber and non-respiratory rat loss (i.e., dead rat) (Fig. 7A), indicating that non-respiratory rat-related losses were minimal.

The rate of n-butyl acetate loss attributable to rat respiratory uptake was well described using the alveolar ventilation rates (estimated from Poet et al., 2003aGo) and a respiratory bioavailability of 100% (fa = 1.0, Fig. 7A). Defined this way, bioavailability is the fraction of material in the inhaled volume (alveolar in this case) of air reaching the blood exchange region in the lung. For all simulations using measured ventilation rates, the cardiac output was set equal to the ventilation rate.

Inhalation-route bioavailability of n-butanol.
The rate of n-butanol chamber-only loss was biphasic (Fig. 7B). Two rate constants were fitted to the chamber-only data, one describing loss during the first phase and one describing loss during the second phase. The fitted rate constants described both chamber only loss and chamber and non-respiratory rat loss (i.e., dead rat) (data not shown), indicating that non-respiratory, rat-related losses were minimal.

The rate of n-butanol loss attributable to rat respiratory uptake was well described using the measured ventilation rates (adjusted to alveolar ventilation rates; Poet et al., 2003bGo) and an average respiratory bioavailability of 50% (FA = 0.5, Fig. 7B). Some variability was apparent: respiratory bioavailability was estimated to be 40% for two rats and 60% for two others.

Validation of Model Predicted Inhalation-Route Blood Kinetics
Blood kinetics following inhalation of n-butyl acetate.
Model predicted blood kinetics of n-butyl acetate were in excellent agreement with the observed blood concentration time-course (Fig. 8A). Predictions of blood n-butanol concentrations were also in relatively good agreement with observed concentrations, though the Tmax was early and Cmax was under predicted by approximately 50%. Simulation of a separate (from the same laboratory, same protocol, Poet et al., 2003aGo) four animal data set showed a poorer fit to blood n-butyl acetate concentrations, with simulations falling outside the 95% confidence limits for all but two points (Fig. 8B). These animals were obtained from a different vendor (see Methods). Model predicted n-butanol blood concentrations were in close agreement with observed data for the first 30 min, thereafter over predicting blood concentrations 50–100%.



View larger version (21K):
[in this window]
[in a new window]
 
FIG. 8. n-Butyl acetate closed chamber inhalation study: model predicted and observed blood concentrations of n-butyl acetate and n-butanol. Average concentrations for a group of two (Panel A, animals 5 and 8, average body weight, 0.303 kg) or four animals (Panel B, animals 1–4, average body weight, 0.35 kg) are shown. Error bars represent ± 2 SE.

 
The difference between the fits for the first and second data sets is likely attributable to experimental variability or observed differences in the two groups of animals. Given two of the known differences between the experiments, ventilation rates somewhat lower and initial chamber concentrations of n-butyl acetate slightly higher (1625 ppm vs. 1400 ppm) in the four animal group (Fig. 8B), the model prediction of blood n-butyl acetate concentrations similar to those in the two animal group (Fig. 8A) is expected because the dose rates would be similar. The experimental data in the four animal group, in contrast, shows n-butyl acetate concentrations that are roughly half of those in the two animal group. Such large differences in experimentally measured blood concentrations would not be expected, and are not attributable to known differences analytical techniques or handling of the blood samples. Differences in body weights alone between these groups are also not likely to be responsible for large differences in measured blood concentrations. Although it cannot be ruled out, a ~15% increase in body weight is unlikely to result in increases in metabolism necessary to reduced blood concentrations by a factor of two. In addition, to the degree principles of allometry hold, the model accounts for body weight related differences in metabolism. However, the difference could be related differences in the physiology of this second group of animals since they appeared to be lethargic.

Blood kinetics following inhalation of n-butanol in rats.
Without adjustment, the butyl series model predicted n-butanol concentrations that were modestly (~40%) less than those observed experimentally (Fig. 9) following inhalation of n-butanol. The simulated and observed Tmaxs were similar, but the former was slightly earlier than the latter. The general shape of the blood n-butanol concentration time-course predicted by the model approximated the observed time-course fairly well. N-butyric acid concentrations were modestly over-predicted, although concentrations were generally low, on the order of the limit of quantitation, for the majority of the study period.

Blood kinetics following intratracheal inhalation of n-butyl acetate in rats.
Blood kinetics of n-butyl acetate and n-butanol are strongly ventilation-rate dependent, but Groth and Fruendt (1991)Go did not report ventilation rates. Consequently, this data set is of limited usefulness for verifying model predictions. Nonetheless, evaluation of the model for general consistency with this data set seemed prudent.

In the absence of measured ventilation rates, simulations were conducted at ventilation rates that varied between 7 and 10 l/h/bw0.75. These ventilation rates were chosen to result in predicted blood n-butyl acetate concentrations in the observed range while remaining in the range of biologically plausible values (Fig. 10A). The presumption here is that n-butyl acetate metabolism is well characterized relative to ventilation rates for this study, and fitting should be accomplished by changing ventilation rates. The resulting ventilation rates were low, as would be expected for anesthetized, intubated rats. Predicted blood n-butanol concentrations were similar to those observed experimentally, but tended towards modest under-prediction (Fig. 10B). Despite the major uncertainties in simulating these data (unknown ventilation rates, experimental details poorly described), the model is reasonably consistent with these data.



View larger version (21K):
[in this window]
[in a new window]
 
FIG. 10. Model predicted and observed blood concentrations of (A) n-butyl acetate and (B) n-butanol during intratracheal inhalation exposure to n-butyl acetate (Groth and Freundt, 1991Go). Ventilation rates (not measured) were varied from 7 (bottom line) to 10 l/h/BW0.75 (top line) to evaluate the model parameterization.

 
Inhalation route blood kinetics of n-butanol in humans.
Model-predicted arterial n-butanol concentrations in male humans following inhalation of 100 ppm n-butanol at various exercise levels (Astrand, 1976Go, Series II) agreed well with observed concentrations (Fig. 11A). Simulations were within 2 SE of the data during exposure. Moreover, the simulated post exposure arterial blood n-butanol time course was not inconsistent with data collected from a single individual (Fig. 11A, post exposure data points). Model simulated arterial blood n-butanol concentrations were approximately 50% higher than observed values at steady state, but generally within 2 SE of the mean values (95% confidence limits on the mean; Fig. 11B) following exposure to 200 ppm n-butanol (Astrand, 1976Go, Series I).

Sensitivity and Uncertainty Analysis
The results of the sensitivity and uncertainty analyses for rat and human PBPK models are presented in Tables 4 and 5, respectively. Analyses supporting the uncertainty designations are presented in Appendix 1. Sensitivity coefficients were determined for all model parameters. Only parameters with sensitivity coefficients greater than 0.1 are reported and evaluated for uncertainty. Parameters with sensitivity designations of H or M and uncertainty designations of H or M are assumed to have sufficient potential to add uncertainty to model simulated arterial n-butanol concentrations to warrant further discussion.


View this table:
[in this window]
[in a new window]
 
TABLE 4 Rat Model Sensitivity and Uncertainty Designations

 

View this table:
[in this window]
[in a new window]
 
TABLE 5 Human Model Sensitivity and Uncertainty Designations

 
For n-butyl acetate exposures in the rat, the key parameters are ventilation rate, other perfused tissues ADH activity, urinary n-butanol elimination rate, and hepatic ADH Vmax (Table 4). The sensitivity coefficients for these parameters following exposure to 3000 ppm n-butyl acetate were 1.1, –0.54, –0.18, and –0.2, respectively. Modeled arterial n-butanol concentrations were only sensitive to hepatic ADH Vmax at high (3000 ppm) n-butyl acetate exposures. Hepatic elimination was blood flow limited at lower concentrations. Although the other parameters (ventilation rate, other perfused tissue ADH activity and urinary n-butanol elimination rate) have uncertainty designations of H or M, they have been used successfully to simulate a large pool of rat toxicokinetic data and are therefore expected to be characterized fairly well.

Only ventilation rate, with a sensitivity coefficient of 1.0, had sufficient sensitivity (H) and uncertainty (M) to influence uncertainty in arterial blood n-butanol in rats following inhalation exposure to n-butanol. Human blood n-butanol concentrations were not sensitive to any parameters with uncertainty designations of M or H (Table 5). It should be pointed out, however, that ventilation rates vary considerably with activity level. Nonetheless, the predictions of arterial n-butanol concentrations during inhalation exposure to n-butanol made with the human PBPK model are generally insensitive to parameters with high or medium levels of uncertainty. Rat blood:air partition coefficients were used as a proxy for human values, though, for other volatile organics, human blood:air partition coefficients are approximately 1.5–2 times lower. Because modeled blood concentrations are insensitive to blood:air partition coefficients (SC < 0.01), this choice has no impact on the results presented here.

Rat and Human Tissue Dosimetry
The rat and human models were used to simulate steady state arterial blood concentrations of butyl series compounds following n-butyl acetate exposure (rats only) and n-butanol exposure. The results are presented in Tables 6 (rat), and 7 (human). Arterial blood concentrations of n-butyl acetate, n-butanol, and n-butyric acid at the rat NOAEL of 500 ppm n-butyl acetate were 20.4 µM, 37 µM, and 40.1 µM, respectively. The corresponding values for the 3000 ppm rat n-butyl acetate NOAEL are 122 µM, 235 µM, and 59 µM.
TABLE 7 Modeled Steady State Human Arterial Blood Concentrations During Inhalation Exposure to n-Butanol



Arterial blood concentrations (µM)

n-Butanol exposure (ppm)

n-Butanol

n-Butyric acid

100 3.9 0.3
200 7.8 0.7
300 11.8 1.0
400 15.7 1.3
500 19.6 1.7
600 23.6 2.0
700 27.5 2.4
800 31.4 2.8
900 35.4 3.1
1000 39.4 3.5
1500 59.2 5.6
2000 79.2 7.9
3000 119.5 13.3
4000 160.5 20.0
5000

202.2

28.3


View this table:
[in this window]
[in a new window]
 
TABLE 6 Modeled Steady State Rat Arterial Blood Concentrations During Inhalation Exposure to n-Butyl Acetate or n-Butanol

 
The use of blood AUC, rather than blood concentrations, as the relevant dose metric for use in calculating HECs has the benefit of direct incorporation of an adjustment for exposure duration. Arterial blood n-butanol AUCs (µM x h, for a week of exposure) were calculated in rats following inhalation exposure to a range of n-butyl acetate and n-butanol exposure concentrations and in humans for a range of n-butanol exposure concentrations (Fig. 12). Rat AUCs were calculated for a study week: 6 h/day, 5 days exposure per week. Human AUCs were calculated assuming constant exposure for a 7-day week. Average weekly AUCs for both rats and humans are the same when calculated for longer exposure durations.

Human Equivalent Concentrations
The HECs for rat NOAELs of 500 ppm (reduced weight gain) and 3000 ppm (neurotoxicity) n-butyl acetate are 169 ppm and 1066 ppm, respectively. While the PBPK models can be used in the future to derive HECs for other rodent n-butyl acetate or n-butanol exposures, an alternative approach was developed that allows non-PBPK model users to derive HECs. Equations describing the relationship between n-butyl acetate (rat only) or n-butanol (human and rat) exposure (Fig. 12) and weekly arterial blood n-butanol AUC were combined and solved for the relationship between rat external exposures and human exposures producing the same weekly arterial blood n-butanol AUC (the HEC). The rats AUCs reflect the 6 h/day, 5 days/week exposure protocol of the toxicity study and the human AUCs are calculated for continuous exposures, thus accounting for the required exposure duration adjustment. The derivation is presented in Appendix 2. These algebraic equations are equivalent to using the PBPK model, and like the model reflect our best understanding of the kinetics of these compounds. The equations can be used with highest confidence in the range of the rat (~2000 ppm n-butyl acetate/butanol) and human (100–200 ppm n-butanol) inhalation exposures used to parameterize the model. In light of the consistency in the kinetics of these compounds at high and low iv doses and the scalability of the model from rat to human, we might expect the approach to be valid across a wider dose range. Given a rat n-butyl acetate NOAEL, the HEC may be calculated using the formula:

Given a rat n-butanol NOAEL, the HEC may be calculated using the formula:

The error in the HEC introduced by using the formulas rather than the PBPK models is less than 1%. The HECs calculated using the first equation are 168 ppm and 1068 ppm and compare well to the model derived values of 169 ppm and 1066 ppm.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
Toxicity is expected to be a more direct function of target tissue exposure than external exposure. Estimating risks associated with human exposure to volatile organic compounds by regression of the exposure-response relationship has specific limitations. Awareness of these limitations has promoted the development of dosimetry based approaches to risk assessment.

In the case of the butyl series compounds, the exposure route of concern for humans is inhalation. Such exposure is expected to result in systemic exposures to both the inhaled compound and its metabolites, though systemic exposure to n-butyraldehyde is believed to be very low because it is not experimentally measurable in the blood. The model addresses this by assuming rapid and almost complete intra-hepatic metabolism of n-butyraldehyde, which appears consistent with the rapid appearance of n-butyric acid in the blood. Inhalation-route subchronic toxicity studies have been conducted for n-butyl acetate in rats (David et al., 1998Go, 2001Go), providing toxicity data attributable to n-butyl acetate, n-butanol, and/or n-butyric acid exposure. The dosimetry-based approach proposed for interpreting the toxicity studies and extrapolating the results to humans involved the development of a PBPK model. Here we have developed and validated a rat PBPK model for inhalation pharmacokinetics of the butyl series compounds n-butyl acetate and n-butanol. The resulting rat model can be used directly to derive RfCs by determining the NOEL or benchmark dose for one butyl series member given toxicity data from a different family member. Standard default procedures can be applied to arrive at the RfC, or the human model can be utilized to conduct the species extrapolation.

Particularly important to the validation of the PBPK model for the inhalation route was a novel combination of experimental approaches—typically used independently—to simultaneously measure ventilation rates, closed chamber loss (respiratory uptake), and blood concentrations of the inhaled volatile and its metabolites (Poet et al., 2003aGo,bGo). The availability of chamber loss data and time-matched ventilation rates allowed direct determination of the respiratory bioavailability of n-butyl acetate and n-butanol. Absent ventilation rates, respiratory bioavailability is estimated based on a presumed ventilation rate for the animals. This approach is particularly problematic when respiratory depression and/or stimulation are observed, as was the case with n-butyl acetate and n-butanol. The respiratory bioavailability of n-butyl acetate was 100% of alveolar ventilation (~60% of minute ventilation) and the respiratory bioavailability of n-butanol was approximately 50% (range 40–60%) of alveolar ventilation (~35% of minute ventilation).

Without adjustment of metabolic parameters, model-predicted and observed blood kinetics of n-butyl acetate, n-butanol, and n-butyric acid following inhalation of n-butyl acetate and n-butanol were, overall, in reasonably close agreement. Successful prediction of inhalation-route blood kinetics from iv kinetic studies implies little or no difference in metabolic production and clearance of the butyl series compounds between these routes of exposure. The modest discrepancies observed, for instance underpredictions of n-butanol concentrations following n-butanol inhalation, could reflect either experimental variability, lower than expected rates of metabolism, or some route-specific difference in n-butanol clearance. This difference would have to be associated with the respiratory tract, and it is not clear why metabolism of n-butanol would be expected to be lower following inhalation exposure. Further, in all the iv kinetic studies, the rate of n-butanol clearance agrees well with the slope of the elimination phase, but in the inhalation kinetic studies, blood n-butanol concentrations appear somewhat low. An explanation consistent with these findings is that the distribution of n-butanol is somewhat different in the inhalation and iv studies. For studies of similar duration, this would be unexpected. However, the iv studies are conducted over a period of minutes, whereas the inhalation study period is on the order of hours. Given the very high rate of loss attributable to the chamber and apparatus observed in the chamber only experiments, it is also possible that variability in this loss rate leads to over-attribution of loss to chamber loss, and under-prediction of uptake of n-butanol. This would also result in under prediction of blood n-butanol concentrations by the model for n-butanol exposures. It should be pointed out, however, that measured chamber loss of n-butanol in the presence of expired rats was consistent across experiments, showing no evidence of high variability. We also evaluated the impact of including respiratory tract metabolism of n-butyl acetate, but found, as expected, that because of the large amount of blood n-butyl acetate metabolism, adding respiratory tract metabolism did not improve fits to blood concentration data.

The close agreement between model-predicted and observed arterial blood kinetics of n-butanol in humans following inhalation of n-butanol imply that most of the processes controlling clearance of n-butanol (urinary elimination, metabolism) scale well using the principles of allometry. More importantly, the available n-butanol toxicokinetic data in humans was sufficient to provide verification of the inhalation route (n-butanol exposure only) for the human PBPK model.

Upper respiratory tract extraction of acetate esters is relatively high (Morris, 1990Go). The absence of an upper respiratory tract in this model is not expected to confound the evaluation of and prediction of blood kinetics of n-butyl acetate and n-butanol. In the case of n-butanol, the model incorporates a bioavailability term that accounts for any "wash-in, wash-out" behavior (Johanson, 1991Go) and yields a net uptake as a function of ventilation rate, pulmonary uptake, and the bioavailability. Further, upper respiratory tract metabolism of n-butanol is not expected to significantly impact uptake due to pre-pulmonary clearance. There is no kinetic evidence of this occurring in our analysis of the experimental data. This would appear as significant over-predictions of n-butanol blood concentrations by the model due to the absence of this route of pre-pulmonary clearance. The fate of n-butyl acetate cleared by the upper respiratory tract is either passage to the blood or metabolism in respiratory or olfactory tissue to n-butanol, which would be passed, perhaps with some additional metabolism, to the blood. The successful use of a simple model structure, absent the upper respiratory tract, to predict n-butyl acetate and n-butanol concentrations implies that lumping upper respiratory tract metabolism (clearance) of n-butyl acetate into representations of respiratory clearance is sufficient. The failure of this approach would be observed as (1) higher rates of vascular metabolism of n-butyl acetate required to fit inhalation versus iv kinetic data and (2) model under-prediction of n-butanol concentrations following inhalation exposure of n-butyl acetate. In our analysis of the kinetic data, there is little evidence of either.

The overall consistency between predicted and observed blood kinetics following inhalation of n-butyl acetate and n-butanol in rats is an acceptable verification of the rat PBPK model's predictive ability related to internal exposures stemming from inhalation exposures. The modest differences observed are on the order of 50%, which is small enough to be attributed to experimental variability. Further, these differences are also small in comparison to the level of uncertainty associated with other aspects of risk assessment for these compounds, and it should in general be viewed favorably that dependence on exposure concentration can be replaced with predicted internal exposures that fall within a factor of two of observed concentrations. The sensitivity of simulated n-butanol concentrations to parameters with medium to high levels of uncertainty was low for the rat model, adding to confidence in its utility as a predictive tool. Similarly, the consistency between model-predicted and observed n-butanol blood concentrations in humans is verification of the human PBPK model for the n-butanol inhalation route. This is the only route of exposure/chemical currently of importance for the human model. The absence of parameters in the human model with a sufficient uncertainty to impact simulated arterial blood concentrations extends confidence in the predictive value of the model. Beyond the range of the rat (~2000 ppm n-butyl acetate/butanol) and human (100–200 ppm n-butanol) inhalation exposures used to parameterize the model, the predictability of the model is not known. Given the consistency in the kinetics of these compounds at high and low iv doses and the scalability of the model from rats to human, we might expect the approach to be valid across a wider dose range, including the range used here for deriving HECs (500–3000 ppm n-butyl acetate for rats, ~100–1000 ppm n-butanol in humans).

The product of this effort, rat and human PBPK models for the butyl series compounds, validated for certain routes of exposure, illustrates the effectiveness of broad multi-institutional public/private collaborations in the pursuit of developing state of the art tools for risk assessment. The collaboration here involved use of the model for designing kinetic studies, including novel closed chamber studies pioneered by Poet et al. (2003aGo,bGo). With the development of these PBPK models for interspecies extrapolation of the internal dosimetry of butyl series compounds, the quantitative tools necessary for conducting scientifically sound, credible risk assessments for these compounds have become available.


    APPENDIX 1
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
Evaluation of Parameter Uncertainty
The evaluation here is based on a limited number of sources and is semi-quantitative at best. A more detailed analysis was out of the scope of this effort, but it is clear that uncertainty analyses of this type would benefit from a more comprehensive analysis.

Minute ventilation.
The analysis here is based on alveolar ventilation rates in Brown et al. (1997)Go. The human value is 5 ml/min/100 g body weight, with a range of 3.5 to 7.7 ml/min/100 g body weight. The rat value is 52.9 ml/min/100 g with a range of 31.5 to 137.6. The range in humans is narrow enough to suggest a CV that is less than 50%. We assign the human uncertainty designation of L and the rat a designation of M.

Cardiac output: From Brown et al., 1997.
The human value is 5200 ml/min. Range given is 4600 to 6500 ml/min. The mean rat value is 110.4 ml/min with a SD of 15.6. This parameter is well known and has a small CV and so its uncertainty determination is low (L) for both species. The principal issue with this parameter is not uncertainty per se, but variability as the level of activity changes.

Hepatic blood flow.
The hepatic blood flow is calculated as the product of cardiac output (uncertainty assessed above) and the fractional hepatic blood flow. The values are from Brown et al. (1997)Go. The fractional hepatic blood flow has a mean of 17.4% (portal plus hepatic arterial flow) and a range of 12–23.6%. This range implies a CV of less than 50%. The mean value for humans is 25% (males) to 27% (females). The range of 11–34% implies a CV of variation of less than 50%. The uncertainty designation is L for humans and rats.

Volume of the OP tissues.
The value of this parameter is calculated as body volume, minus bone volume, minus other tissues represented directly in the model. Because body weight is specified, and tissue volumes are relatively well characterized with small CV, the uncertainty designation is L for rats and humans.

Volume of the venous blood.
The value of this parameter is calculated as body volume, times the fractional volume of the venous blood. The mean value for the rat is 0.045 and the mean value for the human is 0.0514 (Bernareggi and Rowland, 1991Go). Neither ranges nor SD were reported for these values. However, our understanding is that these are known with some accuracy. The uncertainty designation is L for both species.

OP tissues partition coefficient.
The rat value is equal to the value for muscle reported by Kaneko et al. (1994)Go. While the OP tissue compartment is composed of more than muscle, it is dominated by muscle. Further, partition coefficients for this water soluble compound did not vary much between tissues and the CV was smaller than 50% (Kaneko et al., 1994Go). The human value was assumed to be equal to the rat value, which is expected to be a good approximation (Fiserova-Bergerova, 1975Go). The uncertainty designations are low for the rat and low for the human.

OP tissues ADH activity.
The rat value of this first order rate constant was fitted to a large pool of toxicokinetic data and validated against additional data sets. The uncertainty designation is, appropriately, low for the rat. The value for the human is allometrically scaled from the rat value, but was used to successfully simulate the blood kinetics of n-butanol in humans (Astrand et al., 1976Go). The uncertainty designation for the human is medium (M).

n-Butanol renal elimination rate.
The value in the rat is based on limited data (Barton et al., 2000Go) and cannot be verified directly. The human value is allometrically scaled from the rat value and also cannot be verified. The uncertainty designation is high (H) for both rats and humans.

n-Butyl acetate fractional bioavailability.
The value in the rat was fitted to n-butyl acetate closed chamber and blood concentration data. The value was 100% for all simulations. The value for humans was assumed to be the same, but is not used in these simulations. Because of the generally high esterase activity in the respiratory tract and blood of rats and humans, the fractional bioavailability in humans is not expected to differ significantly from the rat value of 100%. It is not expected that the value would be 50 or 150%, for instance. The uncertainty designation is L for both rats and humans.

n-Butanol fractional bioavailability.
The value in the rat was fitted to n-butanol closed chamber data and blood concentration data. The mean value was 50% with a range of 40–60%, suggesting a CV of less than 50%. The human value was used as reported by Astrand et al. (1976)Go, who obtained the value by direct measurement. The uncertainty designations for the rat and human are L.

Hepatic ADH Vmax.
The value for the rat was estimated by fitting a large pool of toxicokinetic data. Generally, the fitting exercises indicated that hepatic metabolism of n-butanol was very high and hepatic clearance was limited by blood flow. The consistency with the large pool of toxicokinetic data gives the appearance of certainty, but it should be pointed out that it is the lower bound on hepatic ADH activity that is identified in this approach. The human value was allometrically scaled from the rat value, and again represents the lower bound of activity in a system that is flow limited. The uncertainty designation for both rats and humans is high (H).

Hepatic ADH Km.
The value for the rat was estimated as reported (Barton et al., 2000Go). The human value was assumed to be the same as the rat. Although the rat value has been used successfully to simulate toxicokinetic data, it is not clear if simulations were sufficiently sensitive to the Km to allow for verification of this parameter. The uncertainty designation is high (H) for both rats and humans.


    APPENDIX 2
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
All human equivalent concentrations (HEC) are human n-butanol concentrations resulting in weekly arterial blood n-butanol AUCs (µM x h) under constant exposure conditions, equivalent to those at the given rat n-butyl acetate or n-butanol exposure under the exposure conditions of the toxicology study: 6 h/day, 5 days/week).

From Figure 12:

and

and

So, by rearrangement, the HEC for rat n-butyl acetate exposures is:

and

The HEC for rat n-butanol exposures, by rearrangement is:


    SUPPLEMENTAL DATA
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
Supplemental data include a description of the gas uptake apparatus, a table containing references and descriptions of the experimental data used herein, a table summarizing data/experimental conditions extracted from Astrand et al. (1976)Go, and ventilation rate data that can be used to reproduce the simulations conducted here. The data can be obtained through the publisher's web site (www.toxsci.oupjournals.org) by searching for the abstract to this article.


    NOTES
 
2 Metabolic scheme: Butyl Acetate–(a)->Butanol–(b)->Butyraldehyde–(c)->Butyric Acid–(d)->Fatty Acids. A = Carboxylesterase, b = alcohol dehydrogenase, c = aldehyde dehydrogenase, d = medium chain acyl-CoA synthetase (mACS). Back

The authors certify that all research involving human subjects was done under full compliance with all government policies and the Helsinki Declaration.

This manuscript has been subjected to review by the National Health and Environmental Effects Research Laboratory of the U.S. EPA and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.


    ACKNOWLEDGMENTS
 
The authors wish to acknowledge the OXO-process panel of the American Chemistry Council for funding the development of the model and the analyses contained herein.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIAL AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX 1
 APPENDIX 2
 SUPPLEMENTAL DATA
 REFERENCES
 
Astrand, I., Ovrum, P., Lindqvist, T., and Hultengren, M. (1976). Exposure to butyl alcohol: Uptake and distribution in man. Scand. J. Work Environ. Health 2, 165–175.[Medline]

Barton, H., Deisinger, P., English, J., Gearhart, J., Faber, W., Tyler, T., Banton, M., Teeguarden, J., and Andersen, M. (2000). Family approach for estimating reference concentrations/doses for series of related organic chemicals. Toxicol. Sci. 54, 251–261.[Abstract/Free Full Text]

Bernareggi, A., and Rowland, M. (1991). Physiologic modeling of cyclosporin kinetics in rat and man. J. Pharmacokinet. Biopharm. 19, 21–50.[CrossRef][ISI][Medline]

Brown, R. P., 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, 407–484.[ISI][Medline]

Corley, R. A., Gies, R. A., and Weitz, K. K. (2000). Evaluation of the Respiratory Bioavailability of Ethyl Acetate, Butanol and Propanol Derivatives Using a Closed, Recirclating (Gas Uptake) Inhalation Exposure System, p. 25. Battelle Memorial Institute, Pacific Northwest Division, Richland, WA.

David, R., Tyler, T., Ouellette, R., Faber, W., Banton, M., Garman, R., Gill, M., and O'Donoghue, J. (1998). Evaluation of subchronic neurotoxicity of n-butyl acetate vapor. NeuroToxicology 19, 809–822.[ISI][Medline]

David, R. M., Tyler, T. R., Ouellette, R., Faber, W. D., and Banton, M. I. (2001). Evaluation of subchronic toxicity of n-butyl acetate vapor. Food Chem Toxicol. 39, 877–886.[CrossRef][ISI][Medline]

Deisinger, P. J., and English, J. C. (2000). The in Vivo Pharmacokinetics of n-Butyl Acetate in Rats after Intravenous Administration, p. 22. Toxicological Sciences Laboratory, Health and Environmental Laboratories, Eastman Kodak Company, Rochester, NY.

Deisinger, P. J., and English, J. C. (2001). Pharmacokinetics of n-Butyl Acetate and Its Metabolites in Rats after Intravenous Administration. Toxicological Sciences Laboratory, Health and Environmental Laboratories, Eastman Kodak Company, Rochester, NY.

Deisinger, P. J., and English, J. C. (2003). Determination of In Vitro Rat Tissue Esterase Activity Toward Ethyl Acetate and the Effect of 1% Tween 20 on the In Vivo Hydrolysis of Ethyl Acetate in Rats after Intravenous Administration (A. C. Council, Ed.). Toxicological Sciences Laboratory, Health and Environmental Laboratories, Eastman Kodak Company, Rochester, NY.

Dietz, F., Rodriguez-Giaxola, M., Traiger, G., Stella, V., and Himmelstein, K. (1981). Pharmacokinetics of 2-butanol and its metabolites in the rat. J. Pharmacokinet. Biopharm. 9, 553–576.[CrossRef][ISI][Medline]

DiVincenzo, G. D., and Hamilton, M. L. (1979). Metabolic fate of [1-14c]isobutyric acid in the rat. Toxicol. Appl. Pharmacol. 47, 609–612.[CrossRef][ISI][Medline]

Fiserova-Bergerova, V. (1975). Biological – mathematical modeling of chronic toxicity. Aerospace Medical Research Laboratory, Wright-Patterson Air Force Base, OH.

Gargas, M. L., Andersen, M. E., and Clewell, H. J., III(1986). A physiologically based simulation approach for determining metabolic constants from gas uptake data. Toxicol. Appl. Pharmacol. 86, 341–352.[CrossRef][ISI][Medline]

Groth, G., and Freundt, K. (1991). Blood alcohol under the presence of n-butyl acetate. Blutalkohol 28, 166–173.[Medline]

Johanson, G. (1991). Modeling of respiratory exchange of polar solvents. Ann. Occup. Hyg. 35, 323–339.[ISI][Medline]

Kaneko, T., Wang, P. Y., and Sato, A. (1994). Partition coefficients of some acetate esters and alcohols in water, blood, olive oil and rat tissues. Occup. Environ. Med. 51, 68–72.[Abstract]

Knights, K. M. (1998). Role of hepatic fatty acid:coenzyme A ligases in the metabolism of xenobiotic carboxylic acids. Clin. Exp. Pharmacol. Physiol. 25, 776–782.[ISI][Medline]

Morris, J. (1990). First-pass metabolism of inspired ethyl acetate in the upper respiratory tracts of the F344 rat and Syrian hamster. Toxicol. Appl. Pharmacol. 102, 331–345.[CrossRef][ISI][Medline]

Poet, T. S., Corley, R. A., and Woodstock, A. (2003a). Evaluation of the Respiratory Bioavailability and Pharmacokinetics of a Series of Esters and Alcohols in Rats. Batelle, Pacific Northwest National Laboratory, Richland, WA.

Poet, T. S., Creim, J. A., Pierce, J. T., Soelberg, J. J., Lee, K. M., and Corley, R. A. (2003b). Evaluation of the Respiratory Bioavailability and Pharmacokinetics of Butyl Acetate using Whole-Body Plethysmography and Gas Uptake Inhalation. Batelle, Pacific Northwest National Laboratory, Richland.





This Article
Abstract
Full Text (PDF)
All Versions of this Article:
85/1/429    most recent
kfi103v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Teeguarden, J. G.
Articles by Clewell, H. J.
PubMed
PubMed Citation
Articles by Teeguarden, J. G.
Articles by Clewell, H. J., III