Evaluation of Potential Toxicity from Co-Exposure to Three CNS ...

Journal of Occupational and Environmental Hygiene, 2: 127–135 ISSN: 1545-9624 print / 1545-9632 online DOI: 10.1080/15459620590916198

Evaluation of Potential Toxicity from Co-Exposure to Three CNS Depressants (Toluene, Ethylbenzene, and Xylene) Under Resting and Working Conditions Using PBPK Modeling James E. Dennison,1 Philip L. Bigelow,2 Moiz M. Mumtaz,3 Melvin E. Andersen,4 Ivan D. Dobrev,5 and Raymond S.H. Yang1 1

Quantitative and Computational Toxicology Group, Center for Environmental Toxicology and Technology, Department of Environmental and Radiological Health Sciences, Colorado State University, Ft. Collins, Colorado 2 Workplace Studies, Institute for Work & Health, Toronto, Ontario 3 Agency for Toxic Substances and Disease Registry, Atlanta, Georgia 4 CIIT Centers for Health Research, Research Triangle Park, North Carolina 5 Sao Vicente-SP, Brazil

Under OSHA and American Conference of Governmen R tal Industrial Hygienists (ACGIH ) guidelines, the mixture formula (unity calculation) provides a method for evaluating exposures to mixtures of chemicals that cause similar toxicities. According to the formula, if exposures are reduced in proportion to the number of chemicals and their respective exposure limits, the overall exposure is acceptable. This approach assumes that responses are additive, which is not the case when pharmacokinetic interactions occur. To determine the validity of the additivity assumption, we performed unity calculations for a variety of exposures to toluene, ethylbenzene, and/or xylene using the concentration of each chemical in blood in the calculation instead of the inhaled concentration. The blood concentrations were predicted using a validated physiologically based pharmacokinetic (PBPK) model to allow exploration of a variety of exposure scenarios. In addition, the Occupational Safety and Health Administration R and ACGIH occupational exposure limits were largely based on studies of humans or animals that were resting during exposure. The PBPK model was also used to determine the increased concentration of chemicals in the blood when employees were exercising or performing manual work. At rest, a modest overexposure occurs due to pharmacokinetic interactions when exposure is equal to levels where a unity calcu R lation is 1.0 based on threshold limit values (TLVs ). Under work load, however, internal exposure was 87% higher than provided by the TLVs. When exposures were controlled by a unity calculation based on permissible exposure limits (PELs), internal exposure was 2.9 and 4.6 times the exposures at the TLVs at rest and workload, respectively. If exposure was equal to PELs outright, internal exposure was 12.5 and 16 times the exposure at the TLVs at rest and workload, respectively. These analyses indicate the importance of (1) selecting appropriate exposure limits, (2) performing unity calculations, and (3) considering the effect of work load on internal doses, and

they illustrate the utility of PBPK modeling in occupational health risk assessment. Keywords

mixture formula, mixtures, PBPK, synergism, unity calculation

Address correspondence to: James E. Dennison, 1690 Campus Delivery, Center for Environmental Toxicology and Technology, Department of Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, CO 80523; e-mail: dennison @colostate.edu.

I

n the workplace, employees are often or perhaps normally exposed to more than one chemical at a time. Coexposure to chemical mixtures can result from products that contain multiple chemicals or from exposure to more than one product, even at different times if there is an overlap between the clearance of one chemical from the body and exposure to a second chemical. If the chemicals cause toxicity via similar mechanisms, the cumulative toxicity from the mixture will be greater than from either of the individual chemicals. Also, if the chemicals interact within the body, their cumulative toxicity will differ from the sum of toxic responses of the individual components of the mixture. Examples of chemicals with similar mechanisms of toxicity exist for many different chemical classes. Specifically, many organic solvents cause central nervous system (CNS) depression after entering the brain and diffusing into

Journal of Occupational and Environmental Hygiene

March 2005

127

membranes of neurons or supporting brain cells. Examples of CNS depressants are thought to include most hydrocarbons, halogenated hydrocarbons, alcohols, ethers, and similar compounds.(1–3) Many organic solvents are also capable of interacting with other organic solvents. Two types of interactions can be considered, pharmacokinetic (PK) or pharmacodynamic (PD). “Pharmacokinetic” refers to the adsorption, distribution, metabolism, and excretion (ADME) of chemicals; “pharmacodynamic” refers to the chemical’s mechanism of action. In another parlance, pharmacokinetics refers to “what the body does to the chemical,” and pharmacodynamic refers to “what the chemical does to the body.” By “PK interaction,” we mean that the tissue dose (i.e., the concentration of a relevant chemical or metabolite in a target tissue) of one chemical is altered by the presence of another chemical. Organic solvents commonly interact with other solvents because they are principally metabolized by a specific enzyme (primarily in the liver), cytochrome P450 2E1(4,5) (CYP 2E1). During simultaneous exposure, if the concentration of the solvent is high enough it can compete with other solvents for metabolism. This competition can, depending on exposure levels, serve to delay the rate of elimination of the solvents and alter the potential toxicity of exposure. The literature is replete with examples of metabolic inhibition between common organic solvents, such as with mixtures of alkyl benzenes,(6–11) chlorinated ethanes and ethylenes,(3,12–14) and other chemicals.(15) However, below some exposure level, the effect becomes insignificant. Current approaches to assessing the risk of mixed exposures generally suggest making the assumption that cumulative effects are additive for the same mechanism of action, unless there are data available indicating that a significant interaction occurs and also supporting a means of quantifying such interactions. While the U.S. Occupational Safety and Health Administration (OSHA) requires that all chemicals with permissible exposure limits (PELs) listed in Table I, SubPart Z of 1910.1000 (General Industry standard) should be included in the assessment of mixture exposures,(16) typical practice is to include chemicals that have the same mechanism of action. At times, this is crudely determined on the basis of target organ or tissue, that is, if two chemicals are liver toxicants, they would be included in a mixture assessment, although a more detailed analysis of the mechanism of action is preferable. The American Conference of Governmental Industrial Hygienists R (ACGIH ) also recommends assessing the mixed exposure to R chemicals that have threshold limit values (TLVs ) based on

TABLE I. TLVs and PELs for Toluene, Ethylbenzene, and Xylene

PEL TLV 128

Toluene (ppm)

Ethylbenzene (ppm)

Xylenes (ppm)

200 50

100 100

100 100

mechanism of action, using a source of information such as the Documentation of the TLVs.(17) Other sources of information are also broadly available. Many health and safety professionals use the mixture formula to evaluate the potential for cumulative overexposure when mixtures are present.(18) Implementing the mixture formula, which is also referred to as the unity calculation and denoted as EM, is a two-step process: (1) identification of chemicals that cause similar kinds of toxic responses, and (2) for those chemicals, calculating the sum of the ratios of the employee exposure level to each chemical to the occupational exposure limit (OEL) for each chemical (Equation 1). A sum of these ratios exceeding unity (1.0) suggests that exposures should be reduced. EM =

Exposurei OELi

(1)

For instance, if exposure to Chemical A is 35 ppm and to Chemical B is 50 ppm, and both chemicals have an OEL of 100 ppm, the EM would be 35/100 + 50/100 = .85, and the cumulative exposure would not appear to be more significant than an exposure to either chemical alone at its OEL. This method is equivalent to the Hazard Index method suggested by the U.S. Environmental Protection Agency (USEPA) for some risk assessment applications.(19,20) By prorating exposures to the OEL, the mixture formula assumes additivity and ignores commonly occurring PK interactions between components. Physiologically based pharmacokinetic (toxicokinetic) models (PBPK or PBTK models) can be used to evaluate these issues. Such PBPK descriptions have been used to describe the ADME of chemicals and chemical mixtures in laboratory animals and humans for many years.(14,21–23) Basically, such models are able to provide descriptions of the fate of specific chemicals in various parts of the body, such as in blood, brain, or liver, by simplifying the structure of the organism based on anatomical and physiological principles; applying measured partition coefficients to determine absorption and distribution; using measured rates of metabolism and excretion; and validating the models with timecourse data for chemicals in available parts of the organisms. Validated PBPK and similar models are now the preferred basis for chemical risk assessment performed or used by the USEPA and other organizations.(24) One of the benefits of such models is that they permit calculation of the concentration of chemicals or their metabolites in the body tissue of choice (limited of course by the structure of the model) and allow calculation of these internal measures of dose under flexible exposure scenarios within the model’s validation range. They can therefore be used to evaluate the impact of other chemicals on the internal dose of a chemical by incorporating relevant interaction mechanisms. This approach has previously been used to assess the impact of PK interactions.(25,26) In previous work, Haddad et al. evaluated the interactions between toluene, ethylbenzene, and xylene mixtures where the concentrations of each chemical


March 2005

were constrained so that the EM was less than unity using the TLVs as the reference OEL.(25,27) In their analysis,(25) the internal doses ranged from 4% to 11% over unity when the EM was at unity, suggesting a slight overexposure could occur. The increased internal dose was caused by a modest inhibition of metabolism of each chemical by the others present in the mixture. However, this analysis was performed under the constraint that the EM was less than 1.0, meaning that the exposure to each chemical was approximately one-third of the TLV. As the PELs are often higher than TLVs (Table I), the potential for interactions is greater. In this study, we extended the previous analysis(25) to exposures that may be allowable in the United States and other locations. Another important issue that arises relates to the ventilation (breathing) rate of the subjects. At high ventilation rates, much more chemical is inspired and available for absorption through the lung. The fraction of blood flow to the liver may decrease as well, limiting metabolic clearance of the chemical. Thus, alveolar air and blood levels of inhaled solvents tend to increase during exercise.(28–30) However, many of the studies of CNS depression in human volunteers or animals that were used to establish TLVs and PELs were conducted at resting conditions, when CNS levels of the chemical would be lower.(31–33) Therefore, TLVs and PELs based on resting conditions may not be adequately protective of workers who are performing manual tasks. This issue was evaluated by determining the increases in internal doses that would be attributable to exercise. METHODS

T

he present study used a unity calculation that was based on internal doses instead of the traditional unity calculation based on external exposure level, thus taking into account nonlinear PK and the influence of PK interactions on cumulative dose. The calculation was based on the pharmacokinetically derived EM, or for short—the EMPK , which was equivalent to the Biological Hazard Index used by Haddad and co-workers(25) (Equation 2): (Ci @EL) EMPK = (2) (Ci @OEL) where Ci is the concentration of a chemical (or a relevant metabolite) in a target tissue at the exposure level (EL) or, for the denominator in the equation, the chemical concentration at the OEL for the single chemical. For this study, the concentration of chemical in venous blood was used as the measure of internal dose, as has been previously done for CNS depressants.(25,27) For rapidly diffusible compounds such as the alkyl benzenes, the chemical concentration in the brain is expected to be closely equilibrated with that in the venous blood, governed by the brain:blood partition coefficient.(1) Since the concentrations in blood and brain are proportional, essentially identical results would be obtained in the EMPK equation whether using chemical concentration in the venous blood or in the brain. For the present simulation, the biomarker of expo-

sure used in the EMPK calculations was the maximum venous blood concentration observed during an 8-hour exposure to the chemical(s). Generally, acute CNS depression is regarded as related to the peak concentration of chemical in the brain. As the exposures were simulated for constant levels, the blood concentration rose throughout the exposure, reaching a quasisteady-state by the end of the period. Between Hour 7 and Hour 8, the concentration of chemicals increased only 1–2%, but the increases during earlier periods were larger. Therefore, the concentration at the end of the exposure period best represented the maximum concentration. For this analysis, mixtures of toluene, ethylbenzene, and xylene were selected based on their common mechanism of action; the availability of validated PBPK models; and evidence that significant PK interactions occur at exposure levels higher than those studied by Haddad et al.(25) Venous blood concentrations for toluene, ethylbenzene, and xylene were obtained using the PBPK model previously published by Tardif et al.(9) This model was a standard PBPK model similar to many models used for other chemicals in the past. The model was initially developed for rats and then modified for male adult humans. It contained four compartments: fat tissue, slowly perfused tissues, richly perfused tissues (which incorporates the brain), and the liver, where all metabolism was based on a single saturable enzyme representing CYP 2E1. The tissues were perfused by the arterial blood that equilibrated with the alveolar concentration of inspired chemical in accordance with the blood:air partition coefficient. Venous blood returning from each tissue compartment was similarly equilibrated with the tissue according to their tissue:blood partition coefficients. Inhibition of metabolism was addressed as competitive inhibition in the saturable metabolism equation. The Tardif et al.(9) model for toluene, ethylbenzene, and m-xylene was initially developed using data for single chemicals and two-chemical mixtures (three binary mixtures.) The model used literature values for physiological parameters and partition coefficients that were measured in vitro. Metabolic parameters were determined using the venous blood data collected during PK experiments in rats at five 30-min intervals after cessation of a 4-hour exposure to 100–200 ppm of toluene, ethylbenzene, and/or m-xylene. The model was validated first by testing it with venous blood data for toluene, ethylbenzene, and xylene after similar exposures to all three chemicals at the same time. Then, the model was scaled to humans by altering the body weight and other physiological and metabolic parameters in accordance with literature values. Model output was then compared with experimental data obtained during controlled human exposures to mixtures of the three chemicals at levels below the TLV. Reasonable agreement was obtained between the human model and the data. This model predicted that metabolic inhibition was not significant during exposure to 20–30 ppm and below, about one-third of the current TLVs for each chemical (see review by ATSDR(34) ). In this range, the additivity assumption held. However, as the exposure increases above this level, inhibition causes a disproportionately higher tissue dose to occur.


March 2005

129

The Tardif et al.(9) model was used in the present study with no modification other than the exposure concentrations and in the exercise simulations, the alveolar ventilation and blood flows. For the “working” subject, the alveolar ventilation and blood flow values were scaled to values reported for “light work” at 50 watts.(35) The parameter values used for the PBPK model are listed in Table II. Many previous studies have indicated that metabolic inhibition occurs between toluene, ethylbenzene, and xylene.(36–38) Most literature describes this interaction as being caused by competitive inhibition of metabolism by CYP 2E1 in the liver.(7,9,10) Toxicological effects of the three xylene isomers are generally considered to be similar,(39) and the PK of o-xylene has been found to be similar to the PK of m-xylene.(40) Therefore, the results of the present evaluation, based on m-xylene, should be similar to results obtained for other isomers of the compound. Model equations were described previously.(9,10) The model was run in Berkeley Madonna simulation software (version 8.0.2a8; Berkeley, Calif.). However, equivalent calculations can be performed in a spreadsheet software platform, as previously described.(41,42)

chemicals at their respective TLVs were 0.63 mg/L (toluene), 2.31 mg/L (ethylbenzene), and 1.77 mg/L (xylene). Thus, the EMPK was calculated: Ce Cx Ct EMPK = + + (3) 0.63 2.31 1.77 Three scenarios were analyzed. Scenario 1 (low exposure) was based on exposure to 16.7, 33.3, and 33.3 ppm of toluene, ethylbenzene, and xylene, respectively. Under this scenario, an EM based on the TLVs would be 1.0 (acceptable exposure under the TLVs.) Scenario 2 (moderate exposure) was based on exposures of 66.7. 33.3, and 33.3 ppm of toluene, ethylbenzene, and xylene, respectively. These levels were determined by dividing the PEL for each chemical by three. Thus, an EM calculation under OSHA regulations would be 1.0 again. Scenario 3 (higher exposures) was based on exposures of 200, 100, and 100 ppm of toluene, ethylbenzene, and xylene, respectively. These values are equal to the PEL for the three chemicals. In addition, the effect of performing light work was evaluated for each scenario.

Exposure Scenarios and Risk Evaluation Calculation of the EMPK was performed using the TLVs as the reference OEL. This means that the TLVs were assumed to be the benchmark for an acceptable level of exposure. The TLV benchmarks were converted into internal doses by using the PBPK model to determine the venous blood level for each chemical during exposure to the TLV. The peak venous blood concentrations determined by the PBPK model for the three

D

TABLE II.

RESULTS uring Scenario 1, the blood levels to all three chemicals rose quickly during the first 2 hours of exposure and then leveled off, although they gradually continued to rise throughout the 8-hour period (Figure 1). Indeed, the timecourse of all three chemicals was generally quite parallel. The blood concentration of toluene was lower in the low exposure scenario, primarily because the inhaled concentration of toluene was lower. However, the inhaled concentration of

Model Parameters

System Parameters Alveolar ventilation Cardiac output A

Resting

A

Compartment Parameters

Light Work

18 L/hr/kgˆ0.75 18 L/hr/kgˆ0.75

40 L/hr/kgˆ0.75 26 L/hr/kgˆ0.75

Fraction of Total Body Mass (%)

Perfusion as Fraction of Cardiac Output, at Rest (%)

Perfusion as Fraction of Cardiac Output, Light Work (%)

19 2.6 5 62

5 26 44 25

7 13 30 50

Fat tissue Liver tissue Richly perfused tissue Slowly perfused tissue Chemical Parameters

Blood: Air PC

Liver: Blood PC

Fat: Blood PC

SP: Blood PC

RP: Blood PC

Vmax A

Km

Ki → T

Ki > EB

Ki → X

Toluene Ethylbenzene Xylene

15.6 28.0 26.4

5.36 2.99 3.44

65.4 55.6 70.4

1.78 .93 1.59

5.36 2.15 3.44

4.8 7.3 5.5

0.55 1.39 .22

— .33 .77

.79 — 1.5

.17 .23 —

Note: PC = partition coefficient; SP = slowly perfused tissues; RP = richly perfused tissues; Vmax = maximum rate of metabolism mg/hr/kgˆ0.75; Km = affinity constant (mg/L); Ki = inhibition constant (mg/L) of the inhibitor indicated in left column for the substrate indicated after →. A Parameter is scaled to body surface area (kgˆ0.75).

130


March 2005

FIGURE 1. Timecourse of toluene, ethylbenzene, and xylene in human blood after exposure to 16.7, 33.3, and 33.3 ppm (one-third of TLV), respectively, at rest and during light work (Exposure Scenario 1). Blood levels rose rapidly during the first 2 hours of exposure to each chemical and then leveled off. Light work caused an increase in blood levels of 70%-100% over the levels when worker was at rest.

xylene and ethylbenzene were the same. The higher blood concentration for ethylbenzene was primarily a consequence of slower metabolism in the liver. With a higher metabolic rate, xylene was cleared from the blood faster. The effect of a light work exercise regimen was incorporated by altering the parameters for ventilation rate and blood flow to values reported in the literature. When the ventilation rate is higher, more chemical enters the lung and is available for uptake. Consequently, the concentration of the chemical in blood reaches higher levels. At the peak concentration in blood, reached at the end of the exposure period, the ratio of the blood concentrations during light work to the blood concentrations at rest were 1.7–2.0 (average 1.9). Thus, the worker performing light work had a cumulative internal dose about 90% higher than the resting worker. These results are similar to those found previously.(29,43) The initial rise and plateau of blood concentrations was similar during exposure to moderate levels of the three chemicals (Figure 2). In this scenario, toluene blood concentrations were higher than those of ethylbenzene and xylene due to the higher exposure concentration of toluene. However, ethylbenzene blood concentrations were still higher than those for xylene, although exposure to these chemicals was equal. As in Scenario 1, light work caused an increase in blood concentration for all three chemicals, but the relative difference was smaller than in Scenario 1. For Scenario 2, the ratio of the blood concentrations of the chemicals under light

work to the blood concentrations at rest was 1.5 – 1.7 (average 1.6). The difference between Scenario 2 (moderate exposure) and Scenario 3 (higher exposure) was an increase in the exposure levels by a factor of three for each chemical. Therefore, the same patterns were found (Figure 3). Again, the blood concentration of toluene was highest, followed by ethylbenzene and then by xylene. The ratio of the blood concentration during light work to the blood concentration at rest was 1.25–1.35 (average 1.3). Using Equation 3, the EMPK was calculated for each scenario (Figure 4). In this figure, the dashed horizontal line represents an EMPK calculation of 1.0. For the low exposures, an EM calculation would be 1.0. At rest, the EMPK calculation indicates that exposure was slightly greater than 1.0. Indeed, the EMPK value of 1.07 is essentially the same result as previously obtained,(25) indicating a modest level of interactions and higher exposure under mixture conditions. However, at light work rates of exercise, the blood concentrations were higher yet, resulting in an EMPK of 1.86. Thus, light work increases the internal dose of the chemicals to about twice what would be allowed under the TLVs. In Scenario 2 (moderate exposure), the concentration of toluene was increased to 66.7 ppm, one-third of the PEL. Thus, if the PELs were the reference OEL and an EM calculation was performed, the moderate exposure levels would be permitted. However, relative to the TLVs, overexposure was indicated


March 2005

131

FIGURE 2. Timecourse of toluene, ethylbenzene, and xylene in human blood after exposure to 66.7, 33.3, and 33.3 ppm (one-third of PEL), respectively, at rest and during light work (Exposure Scenario 2). Similar pharmacokinetic profiles were found for all three chemicals at all exposure levels. Light work increased blood concentrations by about 60% over levels found during rest.

FIGURE 3. Timecourse of toluene, ethylbenzene, and xylene in human blood after exposure to 200, 100, and 100 ppm, respectively, at rest and during light work (Exposure Scenario 3). The blood levels were more than three times the levels in Figure 2 due to additional metabolic inhibition at higher exposure levels. The blood levels at light work were about 30% higher than at rest.

132


March 2005

FIGURE 4. EMPK calculations for all exposure scenarios. EMPK was calculated (Equation 2) using the blood concentrations for single exposures to each chemical as the benchmark. During low exposures, EMPK was modestly over 1.0 during rest but was about 86% above 1.0 during light work. At moderate exposures, EMPK was 2.9 and 4.6 for rest and light work, respectively. At higher exposures, EMPK ranged from 12.5 to 16.

by the internal blood levels of chemicals. At rest, EMPK is 2.9, suggesting that the cumulative blood levels of the three chemicals would be about three times the level permitted by the TLVs. This was partially due to the increased exposure to toluene, augmented by increased metabolic inhibition between the three chemicals. If a worker was similarly exposed but was performing light work, the EMPK increases to about 4.6. At the higher exposure level, the worker was assumed to be exposed to each chemical at its PEL. An EMPK for resting activity levels of 12.5 suggests that exposure, based on internal dose, is much higher than permitted by the TLVs. Moreover, if the worker was performing light work, the EMPK of 16 is higher yet. Simulations for toluene and ethylbenzene mixtures using the PELs as the benchmark instead of TLVs were also performed (data not shown). When toluene and ethylbenzene exposure was 50% of the PEL, the EMPK was 1.16. The increase in the EMPK from Scenario 1 to this one is due to the increase in metabolic inhibition that occurs at higher exposure levels. Also, changing the ratios of mixture components while maintaining the constraints on total exposure did not materially affect the calculations of EMPK . CONCLUSIONS AND RECOMMENDATIONS In this article, we used a PBPK model to calculate the concentration of chemicals in venous blood and then performed unity calculations based on chemical concentrations in blood, thus factoring in the effect of PK interactions. While most health professionals do not have ready access to PBPK models, the magnitude of the overexposures indicated can be illustrative of the consequences of mixture exposure when PK interactions occur or when work activity levels are higher than in the studies that served as the basis for setting the

OEL, which may be the normal case. Each of these factors leads to increases in blood concentrations. While the only available validated PBPK model for these chemicals was based on male physiology, similar results would be expected for females. Health and safety professionals may omit chemicals from a unity calculation for several reasons. First, they may not perform the unity calculation at all. Second, they may refer to a source of information that omits the appropriate mechanism of action. Third, the mechanism of action may not be listed because appropriate toxicity tests have not been performed. Fourth, the effect may not be “critical.” It may occur at a level slightly higher than the effect that drives the OEL, although still contributing to the cumulative effect to some degree. Fifth, the chemical may not have an OEL at all, although it still causes relevant toxicity. Omission of a chemical that contributes directly to the form of toxicity can underrepresent the actual cumulative dose of chemicals to the target tissue and can suggest that the actual exposure is consistent with OELs when it is actually higher. Another potential source of overexposure can occur when a mixture exposure includes a chemical that does not contribute to the same mode of toxicity but may interact with other components anyway. If a mixture included toluene, for example, and another chemical that was not a CNS depressant but was an inhibitor of CYP 2E1, blood toluene levels would increase above the levels permitted by the EM calculation. Indeed, as the inhibitor could also be present at levels up to its OEL, more significant inhibition than found in the present analysis could occur. Thus, it would be appropriate to consider reducing exposure to all components that can have metabolic interactions even if modes of toxicity are different. A third situation could occur when a chemical causes toxicity by the same mode of action but at a level above its OEL. In other words, the OEL for the chemical is based on a critical effect other than the toxicity of concern. The dilemma is that if such chemicals are omitted from the unity calculation, an overexposure could occur. On the other hand, if they are included in the unity calculation using the OEL for the nonrelevant critical effect, EM calculations may indicate overexposure when overexposure is not actually occurring. The most appropriate way to analyze such exposures is to use an effect-specific OEL. In other words, one may estimate what the OEL would be based solely on the effect of concern and use this in a unity calculation. This approach would be preferable to omitting the chemical entirely from the analysis of the cumulative exposure. Other mixtures may exhibit interactions that result in nonadditive cumulative risks. These may include any mixture of chemicals that are primarily metabolized by CYP 2E1. If a given component is not present at a sufficiently large concentration in the mixture, it will not inhibit the metabolism of other components. However, even if a component is present at a low concentration, its metabolism may be inhibited if other components are sufficient in concentration. As many organic solvents are substrates for CYP 2E1 (e.g., carbon tetrachloride,


March 2005

133

styrene, tetrachloroethylene),(44) each may serve as a competitive inhibitor of the enzyme. If they are not included in the unity calculation for a mixture, because their principal toxic effect is not CNS depression, they may have a particularly significant effect on metabolism. Other chemicals are metabolized by other isoforms of the CYP enzyme system and could interact with each other. The use of PBPK modeling to evaluate these types of issues is becoming increasingly important.(45–47) Other toxic effects may also be subject to nonadditive effects. The organophosphate and carbamate pesticides may interact at multiple enzyme levels. Other types of pesticides (e.g., organochlorine) could also interact during metabolism or in their effect on critical biological processes.(48) In some cases, interactions may decrease toxicity when bioactivation steps are the subject of inhibition. Potentially significant interactions also may occur when a chemical can be detoxified by one enzyme but also bioactivated to a more toxic compound, such as a carcinogen, by another enzyme. Inhibition of the enzyme that detoxifies is the equivalent of exposing to a higher level of the parent chemical as more of the chemical is forced through the bioactivating pathway. The significance of interactions for these chemicals can be explored through the in silico toxicology methods used in the present analysis if validated PBPK models exist for particular mixtures of them. The following recommendations arise out of the present analysis. Some of these apply principally to mixtures of toluene, ethylbenzene, and xylenes, although many of them are more broadly applicable.

1. It is important to use a unity calculation when mixture exposures occur with chemicals that have a similar mechanism of action. Internal dose-based approaches are preferable when available, but standard Mixture Formula calculations would minimize significant overexposures. 2. For toluene, ethylbenzene, and xylene mixtures, only slight overexposures appear to occur if the EM is less then 1.0 based on TLV exposures. An additional safety factor of 10–15% can be applied to compensate for PK interactions that occur when exposures are less than the TLVs when all interacting chemicals are included in the unity calculation. For CYP 2E1-metabolized chemicals, this range probably applies to most solvents with OELs in the range of 50–300 ppm. 3. The assessment shown in this article demonstrates that it is critical to include in any unity calculations all components of the mixture that have a similar mechanism of action or can cause PK interactions. Omissions of such components can lead to significant overexposures. 4. Up-to-date OELs should be used to determine the potential for overexposure. However, the analysis of exposure should not be limited to chemicals for which there are OELs. Chemicals for which OELs do not exist should be considered in the same manner. 134

5. Many OELs may be based on resting activity levels. For some chemicals, as with the ones in the present evaluation, a worker performing light work may experience significantly higher absorbed doses. Consideration of the activity level should be a part of the assessment of exposure and the allowable exposure should be adjusted accordingly. 6. Much work remains to be done to clarify the doseresponse relationships of toluene, ethylbenzene, and xylene with respect to CNS depression. This work should include development of PBPK models that include biomarkers, improvement in the assessment of mixture PK (including for other compounds), and extension to include a better quantitative understanding of pharmacodynamic effects. ACKNOWLEDGMENTS

D

etails on the original modeling by Kannan Krishnan at the University of Montreal, assistance with figures by Som Lohitnavy, and manuscript review by Jeanne Nasci is gratefully appreciated. This study was supported in part by a Cooperative Agreement from ATSDR (U61/ATU 881475) and NIEHS Quantitative Toxicology Training Grant (T32 ES07321). REFERENCES 1. Bruckner, J.V., and D.A. Warren: Toxic effects of solvents and vapors. In Casarett and Doull’s Toxicology: The Basic Science of Poisons, 6th Ed. C.D. Klaassen (ed.). New York: McGraw-Hill, 2001. pp. 869–916. 2. Caprino, L., and G.I. Togna: Potential health effects of gasoline and its constituents: A review of current literature (1990–1997) on toxicological data. Environ. Health Perspect. 106:115–125 (1998). 3. Dobrev, I. D., M.E. Andersen, and R.S. Yang: In silico toxicology: Simulating interaction thresholds for human exposure to mixtures of trichloroethylene, tetrachloroethylene, and 1,1,1-trichloroethane. Environ. Health Perspect. 110:1031–1039 (2002). 4. Guengerich, F.P., and T. Shimada: Oxidation of toxic and carcinogenic chemicals by human cytochrome P-450 enzymes. Chem. Res. Toxicol. 4:391–407 (1991). 5. Lof, A., and G. Johanson: Toxicokinetics of organic solvents: A review of modifying factors. Crit. Rev. Toxicol. 28:571–650 (1998). 6. Purcell, K.J., G.H. Cason, M.L. Gargas, M.E. Andersen, and C.C. Travis: In vivo metabolic interactions of benzene and toluene. Toxicol. Lett. 52:141–152 (1990). 7. Tardif, R., S. Lapare, G. Charest-Tardif, J. Brodeur, and K. Krishnan: Physiologically based pharmacokinetic modeling of a mixture of toluene and xylene in humans. Risk Anal. 15:335–342 (1995). 8. Imbriani, M., and S. Ghittori: Effects of ethanol on toluene metabolism in man. G. Ital. Med. Lav. Ergon. 19:177–181 (1997). 9. Tardif, R., G. Charest-Tardif, J. Brodeur, and K. Krishnan: Physiologically based pharmacokinetic modeling of a ternary mixture of alkyl benzenes in rats and humans. Toxicol. Appl. Pharmacol. 144:120–134 (1997). 10. Haddad, S., R. Tardif, G. Charest-Tardif, and K. Krishnan: Physiological modeling of the toxicokinetic interactions in a quaternary mixture of aromatic hydrocarbons. Toxicol. Appl. Pharmacol. 161:249–257 (1999). 11. Thrall, K.D., and T.S. Poet: Determination of biokinetic interactions in chemical mixtures using real-time breath analysis and physiologically


March 2005

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

based pharmacokinetic modeling. J. Toxicol. Environ. Health 59:653–670 (2000). Barton, H.A., J.R. Creech, C.S. Godin, G.M. Randall, and C.S. Seckel: Chloroethylene mixtures: Pharmacokinetic modeling and invitro metabolism of vinyl-chloride, trichloroethylene, and trans-1,2dichloroethylene in rat. Toxicol. Appl. Pharmacol. 130:237–247 (1995). El Masri, H.A., J.D. Tessari, and R.S.H. Yang: Exploration of an interaction threshold for the joint toxicity of trichloroethylene and 1,1dichloroethylene: Utilization of a PBPK model. Arch. Toxicol 70:527–539 (1996). Andersen, M.E., M.L. Gargas, H.J. Clewell, and K.M. Severyn: Quantitative evaluation of the metabolic interactions between trichloroethylene and 1,1-dichloroethylene in vivo using gas uptake methods. Toxicol. Appl. Pharmacol. 89:149–157 (1987). Pelekis, M., and K. Krishnan: Assessing the relevance of rodent data on chemical interactions for health risk assessment purposes: a case study with dichloromethane-toluene mixture. Regul. Toxicol. Pharmacol. 25:79–86 (1997). Occupational Safety and Health Administration (OSHA): “Air Contaminants,” Code of Federal Regulations, Title 29, Part 1910.1000 (2004). American Conference of Governmental Industrial Hygienists R (ACGIH ): Documentation of the TLVs and BEIs with Other Worldwide Occupational Exposure Values. Cincinnati, Ohio: ACGIH, 2003. Klonne, D.R., and G.C. Miller: Occupational exposure limits. In The Occupational Environment—Its Evaluation and Control, First Edition, S.R. DiNardi (ed.). Fairfax, Va.: AIHA Press, 1997. p. 23. Mumtaz, M.M., and R.C. Hertzberg: The status of interactions data in risk assessment of chemical mixtures. In Hazard Assessment of Chemicals, Vol. 8, J. Saxena (ed.): Washington, D.C.: Taylor and Francis, 1993. pp. 47–79. United States Environmental Protection Agency (USEPA): Supplementary guidance for conducting health risk assessment of chemical mixtures (EPA/630/R-00/002). Washington D.C.: USEPA, 2000. Filser, J.G., G. Johanson, W. Kessler, P.E. Kreuzer, P. Stei, C. Baur, and G.A. Csanady: A pharmacokinetic model to describe toxicokinetic interactions between 1,3-butadiene and styrene in rats: Predictions for human exposure. In Butadiene and Styrene: Assessment of Health Hazards, IARC Scientific Publication No. 127, M. Sorsa, K. Peltonen, H. Vainio, and K. Hemminki (eds.). Lyon, France: International Agency for Research on Cancer, 1993. pp. 65–78. Andersen, M.E., H.J. Clewell, M.L. Gargas, F.A. Smith, and R.A. Reitz: Physiologically based pharmacokinetics and the risk assessment process for methylene chloride. Toxicol. Appl. Pharmacol. 87:185–205 (1987). Bois, F.Y., M.T. Smith, and R.C. Spear: Mechanisms of benzene carcinogenesis: Application of a physiological model of benzene pharmacokinetics and metabolism. Toxicol. Lett. 56:283–298 (1991). United States Environmental Protection Agency (USEPA): Proposed guidelines for carcinogen risk assessment (EPA/600/P-92/003C). Washington D.C.: USEPA, 1996. Haddad, S., R. Tardif, C. Viau, and K. Krishnan: A modeling approach to account for toxicokinetic interactions in the calculation of biological hazard index for chemical mixtures. Toxicol. Lett. 108:303–308 (1999). Ogata, M., V. Fiserova-Bergerova, and P.O. Droz: Occupational exposures to mixtures of industrial chemicals. Appl. Occup. Environ. Hyg. 8:609–617 (1993). Haddad, S., M. Beliveau, R. Tardif, and K. Krishnan: A PBPK modeling-based approach to account for interactions in the health risk assessment of chemical mixtures. Toxicol. Sci. 63:125–131 (2001). Lapare, S., R. Tardif, and J. Brodeur: Effect of various exposure scenarios on the biological monitoring of organic solvents in alveolar air. I. Toluene and m-xylene. Int. Arch. Occup. Environ. Health 64:569–580 (1993). Pezzagno, G., M. Imbriani, S. Ghittori, and E. Capodaglio: Urinary concentration, environmental concentration, and respiratory uptake of

30.

31.

32.

33.

34.

35. 36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47. 48.

some solvents: Effect of the work load. Am. Ind. Hyg. Assoc. J. 49:546– 552 (1988). Baelum, J.: Toluene in alveolar air during controlled exposure to constant and to varying concentrations. Int. Arch. Occup. Environ. Health 62:59–64 (1990). Andersen, I., G.R. Lundqvist, L. Molhave, O.F. Pedersen, D.F. Proctor, M. Vaeth, and D.P. Wyon: Human response to controlled levels of toluene in six-hour exposures. Scand. J. Work, Environ. Health 9:405– 418 (1983). Olson, B.A., F. Gamberale, and A. Iregren: Coexposure to toluene and p-xylene in man: Central nervous functions. Br. J. Ind. Med. 42:17–122 (1985). Hastings, L., G.P. Cooper, and W. Burg: Human sensory response to selected petroleum hydrocarbons. In Advances in Modern Environmental Toxicology, Vol. 6, H.N. MacFarland (ed.). Princeton: Princeton Scientific, 1986. pp. 255–270. Agency for Toxic Substances and Disease Registry (ATSDR): Interaction profile for benzene, toluene, ethylbenzene, and xylenes (BTEX). Atlanta: U.S. Department of Health and Human Services, Public Health Service, 2001. Kumagai, S., and I. Matsunaga: Physiologically based pharmacokinetic model for acetone. Occup. Environ. Med. 52:44–352 (1995). Engstrom, K., V. Riihimaki, and A. Laine: Urinary disposition of ethylbenzene and m-xylene in man following separate and combined exposure. Int. Arch. Occup. Environ. Health 54:55–363 (1984). Tardif, R., S. Lapare, K. Krishnan, and J. Brodeur: A descriptive and mechanistic study of the interaction between toluene and xylene in humans. Int. Arch. Occup. Environ. Health 65:135–137 (1993). Mortensen, B., P.O. Osvoll, T. Woldbaek, K. Zahlsen, I. Eide, and O.G. Nilsen: In vitro screening for metabolic interactions among frequently occurring binary mixtures of volatile organic chemicals in Norwegian occupational atmosphere. Pharmacol. Toxicol. 83:49–56 (1998). Office of Environmental Health Hazard Assessment, Pesticide and Environmental Toxicology Section, California Environmental Protection Agency (OEHHA): Public Health Goal for Xylene in Drinking Water. Sacramento, Calif.: OEHHA, 1997. Dennison, J.E., M.E. Andersen, and R.S. Yang: Characterization of the pharmacokinetics of gasoline using PBPK modeling with a complex mixtures chemical lumping approach. Inhal. Toxicol. 15:961–986 (2003). Johanson, G., and P.H. Naslund: Spreadsheet programming a new approach in physiologically based modeling of solvent toxicokinetics. Toxicol. Lett. 41:115–127 (1988). Haddad, S., M. Pelekis, and K. Krishnan: A methodology for solving physiologically based pharmacokinetic models without the use of simulation softwares. Toxicol. Lett. 85:113–126 (1996). Nihlen, A., and G. Johanson: Physiologically based toxicokinetic modeling of inhaled ethyl tertiary-butyl ether in humans. Toxicol. Sci. 51:184–194 (1999). Guengerich, F.P., D.H. Kim, and M. Iwasaki: Role of human cytochrome P-450 IIE1 in the oxidation of many low molecular weight cancer suspects. Chem. Res. Toxicol. 4:168–179 (1991). Paustenbach, D.: Pharmacokinetics and unusual work schedules. In Patty’s Industrial Hygiene, Vol. 3, VI: Law, Regulation, and Management, 5th Ed., R. L. Harris (ed.). New York: John Wiley & Sons, Inc., 2000. Andersen, M.E., and J.E. Dennison: Toxicokinetic models: Where we’ve been and where we need to go! Hum. Ecol. Risk Assess. 8:1375– 1395 (2002). Smith, T.J.: Issues in exposure and dose assessment for epidemiology and risk assessment. Hum. Ecol. Risk Assess. 8:1267–1294 (2002). Andersen, M.E., and J.E. Dennison: Mechanistic approaches for mixture risk assessments—Present capabilities with simple mixtures and future directions. Environ. Toxicol. Pharmacol. 16:1–11 (2004).


March 2005

135