Vol. 86, Issue 6, 1977-1983, June 1999
A physiological model for predicting carboxyhemoglobin
formation from exposure to carbon monoxide in rats
Edgar C.
Kimmel1,
Robert L.
Carpenter2,
James E.
Reboulet1, and
Kenneth R.
Still2
1 Geo-Centers, Inc., and
2 Naval Health Research Center
Detachment Toxicology, Wright-Patterson Air Force Base, Ohio
45433-7903
 |
ABSTRACT |
A time-dependent simulation model, based on the
Coburn-Forster-Kane equation, was written in Advanced Continuous
Simulation Language to predict carboxyhemoglobin (HbCO) formation and
dissociation in F-344 rats during and after exposure to 500 parts/million CO for 1 h. Blood-gas analysis and
CO-oximetry were performed on samples collected during exposure and
off-gassing of CO. Volume displacement plethysmography was used to
measure minute ventilation (
E)
during exposure. CO diffusing capacity in the lung
(DLCO) was also measured. Other model parameters measured in the animals included blood pH, total blood volume, and Hb concentration.
Comparisons between model predictions using values for
E,
DLCO, and
the Haldane coefficient cited in the literature and predictions using measured E,
DLCO, and
calculated Haldane coefficient for individual animals were made.
General model predictions using values for model parameters derived
from the literature agreed with published HbCO values by a factor of
0.987 but failed to simulate experimental data. On average, the general
model overpredicted measured HbCO level by nearly 9%. A specific model
using the means of measured variables predicted HbCO concentration
within a factor of 0.993. When experimentally observed parameter
fluctuations were included, the specific model predictions reflected
experimental effects on HbCO formation.
carbon monoxide exposure; carboxyhemoglobin formation prediction; numerical models; Coburn-Forster-Kane equation
| |
INTRODUCTION |
THE DELETERIOUS EFFECTS OF CO are well known, as is the
principle mechanism of action, which is preferential binding to iron in
the Hb molecule to form carboxyhemoglobin (HbCO). The
result is suppression of O2
transport and, subsequently, cellular respiration. Atmospheric CO is
produced by both biological and industrial processes; however, the most
common source of CO is from incomplete combustion of carbon-based
fuels. Estimations for yearly CO emissions range from 350 to 600 million metric tons. Although natural background concentrations of CO
are relatively low [slightly <20 parts/million (ppm) in urban
areas], the potential for exposure to high concentrations of CO
from numerous sources is great (32). Numerous approaches have been
tried to quantify HbCO production from CO inhalation, and measurement
of HbCO has been used as a biomarker of exposure in victims of smoke
inhalation (15).
Several investigators have developed empirical models of HbCO
formation, with mixed results. The simplest of these are linear models
relating HbCO formation to inhaled CO concentration (12, 24).
Unfortunately, these models are of limited applicability. Other
investigators (14, 26) have developed more complex mathematical functions to relate HbCO formation to CO exposure, which have proven to
be more widely applicable. Coburn and colleagues (8) formulated a
physiological description quantifying CO binding to Hb caused by
exposure to CO in humans, known as the Coburn-Forster-Kane (CFK)
equation. Several models for predicting HbCO formation in small
laboratory animals have been developed (23), many as part of efforts to
investigate the toxicity of combustion atmospheres. Sanders and
colleagues (28, 29) used an empirical model of HbCO formation developed
by Hartzell and colleagues (17) in studies of CO-induced incapacitation
in rats. Although the Hartzell model, an adaptation of the
Peterson-Stewart model (26), consistently predicted equilibrium HbCO
concentration (postmortem) within a few percent, it was not a useful
predictor of incapacitation in the test animals. Their studies suggest
that the rate of HbCO formation, as well as its eventual steady-state
level, may be an important factor in CO-induced incapacitation. This
view has become widely accepted (32).
Exposure atmospheres containing CO generally have other constituents as
well, with combustion atmospheres being a prime example. The most
prevalent, and usually most concentrated, of these other constituents
is CO2. Although
CO2 is often not considered to
have high toxicological potency, it is a well-known stimulator of
ventilation. Therefore, CO2 can
play a significant role in the response to atmospheres containing other
inhalation toxins (16). Unfortunately, empirical models of HbCO
formation generally do not account for the effects of changes in
ventilation that may result from simultaneously breathing CO and
CO2. Well-constructed
physiological models such as the CFK can account for changes in
ventilation and other physiological parameters affecting HbCO
formation. Physiologically based models also have the advantage of
being useful as tools for extrapolating dosimetric and toxicological
findings from laboratory animal species to humans. For example,
Andersen and colleagues (4) have modeled CO production and elimination
from xenobiotic metabolism of methylene chloride as part of the
toxicological assessment of this chemical.
We have developed a model of HbCO formation, based on the CFK
differential equation, capable of accounting for changes in ventilation
as well as other pertinent physiological variables. The model accounts
for inhalation of atmospheres the constituents of which vary as a
function of time. Model performance has been evaluated by comparison of
predictions of HbCO formation with published HbCO concentrations in
rats. These data include HbCO levels resulting from exposure to CO
ranging in concentration from 100 to 4,000 ppm. Model predictions were
also compared with HbCO levels in rats exposed to 500 ppm CO in our
laboratory. We find that the CFK equation is valid over this wide range
of inhaled CO concentrations but that the effect of experimental
manipulations must be taken into account in describing HbCO formation
as a function of time in many experiments.
| |
MATERIALS AND METHODS |
Computational methods.
We used a variant of the basic CFK equation, similar to that of
Benignus and Annau (5), which accounts for blood volume (BV), to
formulate the present model as follows
where
and
CO is
endogenous production of CO (in ml/min), BV is in milliliters,
PICO is
partial pressure of inhaled CO (in Torr), PcO2 is partial
pressure of O2 in capillary
blood, HbO2 is total Hb 
HbCO (in mmol/ml), M is the
Haldane coefficient,
DLCO is diffusing capacity of CO in the lung (in
ml · min
1 · mmHg1),
PB is barometric pressure,
PH2O
is water vapor pressure, and A
is alveolar ventilation (in ml/min).
Model differential equations and algebraic expressions describing HbCO
formation from the inhalation of CO were formulated, and numerical
solutions of these equations were performed by using Advanced
Continuous Simulation Language (ACSL) software (Ageis Research,
Huntsville, AL). Physiological constants used in the model either were
values published in the open literature or were measured in individual
animals. By using the ACSL features that ensure well-behaved numerical
solutions of differential equations with discontinuous forcing
functions, the model was constructed to accept changes in model
parameters either as mathematical expressions or as tabulated data. For
example, the model will accept changing inhaled CO concentrations such
as would occur in many exposure atmospheres. In the model validation
presented here, the available experimental data were from inhalation of
constant CO concentrations. We used this capability to account for
changes in physiological parameters such as reduction of BV due to
blood sampling. In developing our general model, we selected values
cited in the literature for various parameters and optimized these
values by adjusting them, within the range of published normal values,
until there was a best fit for published data sets. This process was
repeated for all parameters, except when measured values were cited by the author [i.e.,
DLCO by
Tyuma et al. (31)]. Then that value was held constant, while
other parameters were optimized by the above method. In this manner, a
best fit was developed for each individual published data set, creating
a range of parameter values. Parameters measured during our experiments
using individual animals were substituted into the general model to
form a specific model, which was exercised to determine the extent to
which experimental data could be reproduced.
Animals.
Thirty-two male F-344 rats (312 ± 15.6 g) were used in this study.
Animals were obtained from a commercial source (Charles River
Laboratories, Raleigh, NC) and were housed in plastic cages over
adsorbent bedding material on a 12-h diurnal cycle. Food and water were
provided ad libitum. Two animals were selected at random and killed.
Both were found to be in normal health by a veterinary pathologist
before the investigation.
Exposures.
Twenty-four animals were exposed, in groups of three, to 499 ± 2.0 ppm CO for 1 h with the use of a 12-port, nose-only exposure chamber
(7). Four animals were exposed to room air. A cannula was implanted in
a femoral artery before exposure, and the animals were allowed to
recover from surgical anesthesia to a lighter depth of anesthesia.
Therefore, the animals were lightly anesthetized during exposure
(urethan 1.5 g/kg). The animals were exposed while restrained in a
combination head-out, volume-displacement plethysmograph/exposure tube
(PET) to allow measurement of ventilation during exposure (19).
Exposure atmospheres were produced by mixing compressed air with 5% CO
from a compressed gas cylinder (Matheson Gas, Twinsburg, OH). Mass flow
controllers (model 840, Sierra Instruments, Monterey, CA) were used for
precision control of the exposure mixture. The exposure-chamber CO
concentration was monitored continuously with a wavelength-specific
nondispersive infrared spectrometer (Infinicon model BINOS 00091, Leybold-Heraeus, Frankfurt, Germany). Four additional animals that were
not exposed or had blood samples drawn were used to establish a
baseline value for
DLCO.
Pulmonary function measurements.
Ventilatory parameters, frequency
(f), tidal volume
(VT), and minute ventilation
(E), were obtained by using
volume-displacement plethysmography. Respiratory flow was measured as
pressure change across a screen pneumotachograph in the PET wall by
using a differential pressure transducer (model DP 45-14, Validyne
Engineering, Northridge, CA). The transducer signals were preamplified,
and VT was obtained by
integrating the flow signal (model XA, Buxco Electronics, Sharon, CT).
Ventilation was monitored continuously during the 1-h exposure and for
1 h postexposure. A minimum of 30 breaths/min was measured. After
exposure, the animals were given an additional hour to off-gas CO. At
this time, a tracheal cannula was inserted for determination of
DLCO by
using the plethysmographic (Diamond box, Buxco Electronics) and gas
chromatographic (model 8A, Shimadzu, Tokyo, Japan) method of Kimmel and
Diamond (21).
Blood chemistry and blood-gas analysis.
Cannulas placed in the femoral artery of the exposed animals to allow
real-time sampling of blood were exteriorized through fittings in the
wall of the PETs. Two 1-ml blood samples were withdrawn from each
animal, with the exception of four animals selected at random from
which only one blood sample was drawn. When two samples were taken, the
first sample was taken during CO exposure and the second 1 h later at a
corresponding time point during the 1-h postexposure off-gassing period
(see Table 1). Blood Hb and gas analyses
were performed immediately after the samples were collected (models 682 and 1620, Instrumentation Laboratories, Lexington, MA).
| |
RESULTS |
Model predictions and published data.
The general model for simulating HbCO formation used values for
pertinent physiological parameters obtained from published data (see
Table 2). HbCO formation curves predicted
with this model were compared with published HbCO formation data from
several studies. General model parameters selected from the literature were optimized as previously described. Optimized parameter values were
well within the range of these values reported for rats. As shown in
Figs. 1- 4, model
predictions using these optimized parameter values were fit to HbCO
formation data reported for 20- to 240-min exposures to CO ranging in
concentration from 100 to 4,000 ppm (2, 5, 30, 31). In the case of the
data taken from Andersen et al. (3), the data points were derived from
their model predictions. In data sets for which specific parameter
values were determined, these values were substituted into the general
model, as opposed to the optimized value. For example, where Silbaugh
and Horvath (30) reported VT and
f for their animals (2.5-3.0 ml
and 140 breaths/min, respectively), these values were substituted into
the general simulation model. Similarly, Tyuma and colleagues (31)
reported a
DLCO of
0.13 ml · min1 · mmHg1,
which was used for this simulation. In several instances, the general
model parameters selected differed significantly from those employed by
other investigators. For example, the Hb concentration used for the
general model was 16.2 g/100 ml and was nearly two orders of magnitude
lower than that reported by Benignus and Annau (5). Given the success
of their simulations, we suspect that these investigators did not use
the reported value of 15.8 g/ml in their calculations. Similar to
Andersen and colleagues (3), the present model used a conversion factor
to express M in terms of solution
concentration instead of partial pressures. All parameter values used
in the general model either were means compiled from a variety of
published sources or were derived from well-established allometric
relationships (1-3, 5, 6, 9, 18, 21, 27, 30, 31).
View larger version (10K):
[in this window]
[in a new window]
|
Fig. 1.
General model simulation compared with data from Andersen et al. (3).
HbCO, carboxyhemoglobin; ppm, parts/million.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2.
General model simulation compared with data from Benignus and Annau
(5). Symbols, individual data points from repeated measures at
designated concentrations.
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 3.
General model simulation compared with data from Tyuma et al. (31).
Symbols, individual data points from repeated measures at designated
concentrations.
|
|
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4.
General model simulation compared with data from Silbaugh and Horvath
(30). Symbols, individual data points from repeated measures at
designated concentrations.
|
|
Model predictions and experimental data.
The specific simulation model for HbCO formation used mean
physiological parameters measured in our experimental animals. Some of
these parameters were estimated from measured values by using
well-known physiological relationships or regressions derived from
published data (see Table 3). The time
dependency for a calculated parameter was considered to be identical
with that of the underlying variable. For example, time-dependent
changes in M were considered to be the
same as those for pH. Measured or estimated parameters included
body weight, total BV, total Hb, Hb concentration, blood
O2 partial pressure
[arterial PO2 (PaO2)], blood
CO2 partial pressure
[arterial PCO2
(PaCO2)], DLCO (both
before and after blood sampling; see
DISCUSSION),
f, and
VT. Estimates of
M (11) were derived from blood pH.
Data from Allen and Root (1) were fit by a nonlinear, Lorentzian regression (r2 = 0.90) to develop an empirical expression relating blood pH to
M (see Fig.
5). BV (in ml) was determined for each
animal as a function of body weight by multiplying body weight (in g)
by a factor of 0.0641 (2). The
A was
estimated by multiplying the calculated E
by 0.67 (6).
PcO2 was
determined from measured values of PaO2
by the method of Dickenson (10).
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 5.
Nonlinear least squares fit of Haldane coefficient
(M) vs. pH. Data were taken from
Allen and Root (1).
|
|
Individual simulations were performed by using parameters specific to
that animal. Predicted vs. measured HbCO levels were plotted for both
general (see Fig. 6) and specific model
simulations (see Fig. 7). The slopes of the
linear regressions between predicted and observed HbCO data were used
as estimates of overall model performance. Over the range of HbCO
concentrations examined, the general model-predicted %HbCO was 1.087 of the corresponding observed %HbCO. The specific model-predicted
%HbCO was 0.993 of the corresponding observed %HbCO. The general
model fit the observed data more poorly than did the specific model,
with coefficients of determination of 0.883 and 0.998, respectively. As
shown in Fig. 8, the general model
overpredicted HbCO formation and dissociation in our experimental animals. The slope of the regression for the general model fit suggests
that, when the parameter values selected by the optimization process
described above are used, this model overpredicts HbCO level by an
average of ~9%. Benignus and Annau (5) noted that their predictions
were consistently 7% low and attributed this, in part, to selection of
parameter values,
A in
particular. Similarly, Andersen and colleagues (3) noted that an
adjustment factor of 1.2 was needed to bring predicted and measured
HbCO levels into agreement, particularly during the dissociation phase. In addition, these latter investigators reported using parameter values
that were twice that reported in the literature to achieve agreement
between measured and predicted HbCO formation. The slope of the
regression for the specific model fit (0.993) suggests that this model
underpredicts HbCO by an average of 0.7%, indicating the degree to
which using measured parameter values improves model performance.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6.
Linear least squares fit of general model predicted vs. measured HbCO.
Model parameters derived from best fit of published values.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 7.
Linear least squares fit of specific model predicted vs. measured HbCO.
Model parameters were derived from individual animal measurements.
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 8.
Comparison of specific and general model simulations at 500 ppm CO.
Symbols, individual data points from repeated measures at designated
concentrations.
|
|
During the course of exposure, several physiological changes were
observed in the experimental animals that impacted model performance.
Many of the observed physiological changes could be attributed to the
effects of CO exposure. For example, exposure to CO was found to reduce
blood pH in a dose-response manner. Blood pH in unexposed animals was
7.436 ± 0.007, whereas, after 30 and 60 min of exposure, the blood
pH dropped to 7.316 ± 0.014 and 7.281 ± 0.033, respectively.
Blood pH 1 h after exposure cessation remained low at 7.221 ± 0.046. As pH diminished, there were corresponding changes in
M, with a normal
M of 173 ± 6.1 falling to 164 ± 6.3, 154 ± 12.1, and 143 ± 3.8 at 30, 60, and
120 min, respectively. The PaO2 and,
consequently, the
PcO2 were
elevated slightly by exposure. Normal
PaO2 was 96 ± 13.1 Torr and
gradually increased to 107 ± 10.1 Torr at the end of the
experiment. The PaCO2 in naive animals
was 41.9 ± 0.78 Torr and increased to 49.2 ± 1.71 and 55.8 ± 2.3 Torr at 30 and 60 min, respectively, during exposure. Many of
the observed changes appeared to be related to experimental procedures.
There was a reduction in
DLCO from
successive blood sampling. Drawing a single blood sample reduced
DLCO from
0.15 ± 0.017 to 0.10 ± 0.021 ml · min1 · mmHg1.
A second sample collection further reduced
DLCO to
0.07 ± 0.026 ml · min1 · mmHg1. Depression of
ventilation was observed over the course of the exposure and
elimination period. Average E at the
beginning of the exposure was 147.8 ml/min but dropped to 66.8 ml/min
at the end of the exposure and remained depressed (severely in some animals) at average E of 57.3 ml/min at
60 min postexposure. A similar depression of ventilation was shown in
animals exposed to room air only.
| |
DISCUSSION |
Physiologically, HbCO formation is influenced by a number of
cardiopulmonary factors, such as
DLCO,
A, and
PcO2. We found it necessary to take into account changes in these underlying physiological variables to adequately model HbCO formation and dissociation from CO inhalation. The reduction of BV due to sampling apparently decreased
DLCO. As
noted by Coburn et al. (8), nonuniformity of
DLCO to
A will also
influence mean alveolar CO tension and, therefore, HbCO formation rate.
It is possible that the reduction of BV led to such a maldistribution
of DLCO and
A. Reduced
delivery of O2 to tissues caused
by CO exposure and subsequent HbCO formation has been shown to cause an
elevation in heart rate, which is accompanied by vasodilation, and
reduction of systemic blood pressure (25). Both effects accommodate the
need for greater blood flow to vital organs but could lead to further
maldistribution of
DLCO and
A.
Studies have demonstrated that CO exposure leads to sufficient
disruption of acid-base balance to cause a change in
M (1) and elevation of
PcO2.
Disruption of blood acid-base balance can lead to an elevation of
ventilation (5, 30). In the present investigation, we found a moderate
exposure-related depression of ventilation, despite a reduction in
blood pH, and a moderate elevation of arterial
O2 tension and, consequently,
calculated PcO2. Arterial
CO2 also increased over the course
of the experiment. Given a diminished
DLCO, the
elevation of PaO2
observed could be attributed to a decreased metabolic
rate. However, the observed increase in
PaCO2 would suggest that metabolic rate
did not decrease. Because the buildup of
CO2 in the blood is also
consistent with reduced
DLCO, the
elevated blood O2 tension was most
likely a result of displacement of
O2 from Hb during the formation of HbCO.
The reduction in ventilation that was observed in our animals could be
a result of CO inhalation, which has been shown to decrease ventilation
in humans (13). These changes in ventilation can also be attributed to
experimental conditions not related to CO exposure. Mauderly (22)
demonstrated a depression of both VT and
f in rats housed in PETs. Silbaugh and
Horvath (30) attributed, in part, the cardiovascular and
cardiopulmonary changes they observed in their animals to the effects
of restraint in PETs.
Coburn et al. (8) discussed the sensitivity of the HbCO rate equation
to changes in physiological parameters
(A,
DLCO, and
PcO2). They
noted that a steady state actually never exists in humans, and this is
probably true in experimental animals. However, they point out that
HbCO levels do not vary rapidly, suggesting that the processes
governing body CO burden are highly dampened. They demonstrate that
this is consistent with the mathematical properties of the CFK
equation. Figure 8 also demonstrates this behavior. The general model,
which does not contain any alteration of the physiological parameters,
overshoots the experimental data during CO uptake and undershoots the
observed HbCO levels during dissociation. The specific model, in which
the physiological parameters are changed to reflect the time course of
measured values, follows the observed HbCO data closely. Figure
8 also demonstrates the specific model's ability to respond to
transient (short timescale) changes in physiological parameters.
Although we have not written the set of partial differential equations
that fully describes the rate relationships between HbCO kinetics and
the governing physiological parameters, the numerical procedure used to
solve the CFK equation provides a good correlation between observed data and these model calculations when changes in these parameters are
introduced into the simulation. This approach to solution of the CFK
equation allows the examination of HbCO kinetics on a timescale of
minutes as well as hours. We assume that the general model's inability
to fit uptake-clearance experiments is due to the fact that in such
experiments several physiological parameters are changing
simultaneously, because of inhalation of CO and experimental stress,
whereas model parameters are fixed. Thus numerical solutions to the CFK
equation with time-varying physiological parameters may prove a useful
adjunct to time-course experiments. The specific model fit to the
experimental data in Figure 8 required inclusion of all observed
physiological changes. Thus a numerical solution to such an extended
CFK equation can be used to derive half times for specific exposure and
physiological situations. Comparison of the general and specific model
fit with experimental observations suggests an improved understanding
of the impact of the underlying physiology on HbCO kinetics. This
understanding translates to improvements in assessment of the health
risk associated with CO inhalation, particularly in situations in which
other environmental factors combine to affect the overall physiological
status and, therefore, the response to CO.
| |
ACKNOWLEDGEMENTS |
The authors acknowledge the contributions of Greg Whitehead, Sue
Prues, Petty Officer 1st class Anson Walsh, and Dr. Eldon Smith to this research.
| |
FOOTNOTES |
The Naval Medical Research and Development Command sponsored this
research under Work Unit # 63706N-M00095.004.1714.
The experiments reported herein were conducted according to the
principles set forth in the Guide for the Care and Use
of Laboratory Animals, prepared by the Committee on
Care and Use of Laboratory Animals of the Institute of Laboratory
Animal Research, National Research Council, Dept. of Health and Human
Services (National Institutes of Health Publication No. 85-23,
revised 1985) and the Animal Welfare Act of 1966, as amended.
The opinions herein are those of the authors and are not to be
construed as official or reflecting the views of the Navy Department or
the Naval Service at large.
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: E. C. Kimmel,
Naval Health Research Center Detachment Toxicology, Bldg. 433, 2612 Fifth St., WPAFB, OH 45433-7903.
Received 18 May 1998; accepted in final form 17 February 1999.
| |
REFERENCES |
1.
Allen, T. A.,
and
W. S. Root.
Partition of carbon monoxide and oxygen between air and whole blood of rats, dogs and men as affected by plasma pH.
J. Appl. Physiol.
10:
186-190,
1957[Abstract/Free Full Text].
2.
Altman, P. L.,
and
D. S. Dittmer
(Editors).
Biology Data Book (2nd ed.). Bethesda, MD: FASEB, 1974, vol. III, p. 1413-2123.
3.
Andersen, M. E.,
H. J. Clewell,
M. L. Gargas,
M. G. MacNaughton,
R. H. Reitz,
R. J. Nolan,
and
M. J. McKenna.
Physiologically based pharmacokinetic modeling with dichloromethane, its metabolite, carbon monoxide, and blood carboxyhemoglobin in rats and humans.
Toxicol. Appl. Pharmacol.
108:
14-27,
1991[Medline].
4.
Andersen, M. E.,
H. J. Clewell,
M. L. Gargas,
F. A. Smith,
and
R. H. Reitz.
Physiologically based pharmacokinetics and the risk assessment process for methylene chloride.
Toxicol. Appl. Pharmacol.
87:
185-205,
1987[Medline].
5.
Benignus, V. A.,
and
Z. Annau.
Carboxyhemoglobin formation due to carbon monoxide exposure in rats.
Toxicol. Appl. Pharmacol.
128:
151-157,
1994[Medline].
6.
Brown, R. P.,
M. D. Delp,
S. L. Lindstedt,
L. R. Rhomberg,
and
R. P. Beliles.
Physiological parameter values for physiologically based pharmacokinetic models.
Toxicol. Industr. Health
13:
407-484,
1997[Medline].
7.
Cannon, W. C.,
E. F. Blanton,
and
K. E. McDonald.
The flow past chamber: an improved nose-only exposure system for rodents.
Am. Ind. Hyg. Assoc. J.
44:
923-928,
1983[Medline].
8.
Coburn, R. F.,
R. E. Forster,
and
P. B. Kane.
Considerations of the physiological variables that determine the blood carboxyhemoglobin concentration in man.
J. Clin. Invest.
44:
1899-1910,
1965.
9.
Costa, D. L.,
J. S. Tepper,
and
J. A. Raub.
Interpretations and limitations of pulmonary function testing in small laboratory animals.
In: Treatise on Pulmonary Toxicology: Comparative Biology of the Normal Lung, edited by R. A. Parent. Boca Raton, FL: CRC, 1992, vol. I, p. 367-399.
10.
Dickenson, C. J.
A Computer Model of Human Respiration. Baltimore, MD: University Park Press, 1977, p. 1-18.
11.
Douglass, C. G.,
J. S. Haldane,
and
J. B. S. Haldane.
The laws of combination of haemoglobin with carbon monoxide and oxygen.
J. Physiol. (Lond.)
44:
275-304,
1912.
12.
Forbes, W. H.,
F. Sargent,
and
F. J. W. Roughton.
The rate of carbon monoxide uptake in normal man.
Am. J. Physiol.
143:
594-608,
1945.
13.
Haldane, J. B. S.
The action of carbonic oxide on man.
J. Physiol. (Lond.)
18:
430-462,
1895.
14.
Hanks, T. G.,
and
R. D. Farquhar.
Analysis of Human Performance Capabilities as a Function of Exposure to Carbon Monoxide. Geneva: World Health Organization, 1969. (SystMed Rep. R 9001)
15.
Haponik, E. F.
Clinical and functional assessment.
In: Respiratory Injury: Smoke Inhalation and Burns, edited by E. F. Haponik,
and A. M. Munster. New York: McGraw-Hill, 1990, p. 137-178.
16.
Hartzell, G. E.
Prediction of the toxic effects of fire effluents.
Fire Sci.
7:
179-193,
1989.
17.
Hartzell, G. E.,
H. W. Stacy,
W. G. Switzer,
D. N. Priest,
and
S. C. Packham.
Modeling of toxicological effects of fire gases. IV. Intoxication of rats by carbon monoxide in the presence of an irritant.
In: Advances in Combustion Toxicology, edited by G. E. Hartzell. Lancaster, PA: Technomic, 1989, vol. 2, p. 1-17.
18.
Johnson, J. D.,
D. L. Wetmore,
C. W. Martinez,
and
C. R. Ostrander.
Developmental changes in bilirubin production in the rat.
J. Pediatr. Gastroenterol. Nutr.
2:
142-151,
1983[Medline].
19.
Kimmel, E. C.
A small animal plethysmograph/exposure tube for determination of respiratory mechanics during exposure, using non-invasive methods to measure intrapleural pressure (Abstract).
Toxicol. Sci.
48:
116,
1999.
20.
Kimmel, E. C.,
R. L. Carpenter,
E. A. Smith,
J. E. Reboulet,
and
B. H. Black.
Physiologic models for comparison of inhalation risk between laboratory and field generated atmospheres of a dry powder fire suppressant.
Inhal. Toxicol.
10:
905-922,
1998.
21.
Kimmel, E. C.,
and
L. Diamond.
The role of nicotine in the pathogenesis of pulmonary emphysema.
Am. Rev. Respir. Dis.
129:
112-117,
1984[Medline].
22.
Mauderly, J. L.
Respiration of the F344 rat in nose-only inhalation exposure tubes.
J. Appl. Toxicol.
6:
25-30,
1986[Medline].
23.
Montgomery, M. R.,
and
R. J. Rubin.
The effect of carbon monoxide inhalation on in vivo drug metabolism in the rat.
J. Pharmacol. Exp. Ther.
179:
465-473,
1971[Abstract/Free Full Text].
24.
Pace, N.,
W. V. Consolazio,
W. A. White, Jr.,
and
A. R. Behnke.
Formulation of principal factors affecting the rate of uptake of carbon monoxide by man.
Am. J. Physiol.
147:
352-359,
1945.
25.
Penney, D. G.
Hemodynamic response to carbon monoxide.
Environ. Health Perspect.
77:
121-130,
1988[Medline].
26.
Peterson, J. E.,
and
R. D. Stewart.
Adsorption and elimination of carbon monoxide by inactive young men.
Arch. Environ. Health
21:
165-171,
1970[Medline].
27.
Sabo, J.,
E. C. Kimmel,
and
L. Diamond.
The effects of the Clara cell toxin, 4-ipomeanol, on pulmonary function in rats.
J. Appl. Physiol.
54:
337-344,
1983[Abstract/Free Full Text].
28.
Sanders, D. C.,
and
B. R. Endecott.
The effect of elevated temperature on carbon monoxide-induced incapacitation.
J. Fire Sci.
9:
297-310,
1991.
29.
Sanders, D. C.,
B. R. Endecott,
R. M. Ritter,
and
A. K. Chaturvedi.
Variations of Time-to-Incapacitation and Carboxyhemoglobin Values in Rats Exposed to Two Carbon Monoxide Concentrations. Washington, DC: Dept. of Transportation, FAA, Office of Aviation Medicine, 1993, p. 1-15, A1-A5. (Dept. of Transportation Technical Rep. DOT/FAA/AM-93/7)
30.
Silbaugh, S. A.,
and
S. M. Horvath.
Effect of acute carbon monoxide exposure on cardiopulmonary function in the awake rat.
Toxicol. Appl. Pharmacol.
66:
376-382,
1982[Medline].
31.
Tyuma, I.,
Y. Ueda,
K. Imaizumi,
and
H. Kosaka.
Prediction of carboxyhemoglobin levels during and after carbon monoxide exposures in various animal species.
Jpn. J. Physiol.
31:
131-143,
1981[Medline].
32.
World Health Organization.
Carbon Monoxide. Geneva: World Health Organization, 1979. (Environmental Health Criteria 13 Ser.)
J APPL PHYSIOL 86(6):1977-1983
TOP
ABSTRACT
INTRODUCTION
