Sources and Types of Data for Establishing Spacecraft Water Exposure Guidelines
IN this chapter, the Subcommittee on Spacecraft Water Exposure Guidelines describes the sources and types of data that should be used for establishing spacecraft water exposure guidelines (SWEGs). This information is similar to that described in the National Research Council's (NRC 1992) Guidelines for Developing Spacecraft Maximum Allowable Concentrations for Space Station Contaminants (SMACs), but with important differences. First, the major route of exposure considered in SMACs is inhalation, whereas the major expected route of exposure considered in SWEGs is oral, via ingestion of drinking water and food prepared in potable water. Dermal absorption and inhalation could be secondary routes of exposure. Second, the SWEGs should incorporate advances in the understanding of human physiology and metabolism in microgravity that have occurred since the SMACs were published. Third, the duration of the exposures of interest is different for contaminants in drinking water, because exposure can be avoided if necessary by not drinking the water from the potable drinking-water supply of the spacecraft (emergency supplies could be made available for short periods). It is much more difficult to avoid contact with the ambient air in a spacecraft.
As noted in Chapter 2, contamination of the water used for consumption and personal hygiene could result either from the release of toxic substance into the atmosphere with subsequent absorption into water (e.g., into water condensate collected for reclamation) or because of a malfunction of the water purification system.
Several types of information are evaluated to develop risk-based SWEGs for water, including data on the physical and chemical properties of the substance; human clinical and epidemiologic studies; animal toxicity data from studies, in which exposure ranges from acute to chronic, to identify toxic and carcinogenic end points; in vitro toxicity studies; and mechanistic studies. Data from oral exposure studies are the most relevant for deriving SWEGs, but dermal absorption and inhalation studies should also be evaluated because exposure from those routes occurs when there are volatile constituents in the water or when water is used for hygiene purposes. In addition, those types of data can be used to predict oral toxicity, in the absence of relevant oral data, when toxicokinetic and metabolic data are available to predict disposition and toxicity after oral administration.
CHEMICAL AND PHYSICAL CHARACTERISTICS OF A TOXICANT
The chemical and physical characteristics of a chemical influence its absorption, distribution, metabolism, and excretion from the body. When they are expressed quantitatively, the effect is called the pharmacokinetics (or toxicokinetics) of the compound. For example, molecular size, stability in stomach acid, water solubility, and overall lipophilicity of ingested substances strongly influence their ability to be absorbed and their subsequent pharmacokinetics. More information is needed to determine whether microgravity or other factors alter the physiologic disposition of toxic substances based on their chemical and physical properties.
Ideally, SWEGs should be established for all compounds that might be found in water onboard spacecraft. As a practical matter, it would be difficult and costly to develop SWEGs for the more than 400 chemical species identified to date in the Mir space-station water system. So, the National Aeronautics and Space Administration (NASA) must identify the compounds that cause the most concern and rank them
accordingly for SWEG development. Those criteria are discussed in Chapter 5.
Human toxicity data frequently are obtained from epidemiologic studies of low-level, long-term environmental and industrial exposures as well as short-term exposures, usually to large quantities of toxicants after accidents. These data sometimes provide a basis for estimating a dose-response relationship.
However, epidemiologic studies have several limitations. Most studies involve retrospective analyses that provide only a small amount of data on past chemical exposures (Checkoway et al. 1989). For example, in occupational studies, exposures are typically estimated from records based on the number of years of employment in a given place or from available records on the work environment or personal samples (Rinsky 1989). Prospective studies, which involve assessment of exposure while the cohort is followed, can provide more reliable information on exposure because the sampling scheme can be devised as part of the research plan rather than relying on available data collected for other purposes, such as assessing compliance with exposure regulations (Smith 1987). However, these studies also have limitations because some outcomes of exposure are not manifested for months or even years after exposure.
Epidemiologic studies also vary in the accuracy and precision of the health outcome measured. Some outcomes must rely on information collected for other purposes, such as death certificates, which can show an error rate of 20-40% in the United States (Percy et al. 1981); early preclinical testing; pathology reports; or early preclinical markers of pathology. Despite these limitations, if the populations studied are large enough, have had substantial exposure, and have had a sufficient interval between exposure and study to allow for the expression of disease, epidemiologic studies offer the advantage of providing human data. Epidemiologic studies often can provide data that assist in establishing a permissible concentration for human exposure (e.g., Threshold Limit Values of the American Conference of Governmental Industrial Hygienists).
Epidemiologic outcomes often are reported in terms of relative risk,
a ratio of the rate of outcome of disease or disability in the exposed population to that in the nonexposed population. Care must be taken in using this type of information because relative risk is not a measurement of risk. For example, the relative risk for a rare disease can be the same as that for a common disease and lead to substantially less total risk. To determine an acceptable exposure level, information on relative risk should be analyzed to identify the relationship of relative risk and risk of morbidity or mortality. The risk that is acceptable is a matter of policy and could vary significantly depending on which population is exposed.
Most risk assessments use data derived from laboratory animal studies combined with human clinical and epidemiologic data, when available. To assess either the acute or the chronic toxicity of a water contaminant, emphasis should be placed on human data, provided that it is sufficiently reliable. Using data directly from humans obviates the need to estimate relative sensitivities of humans and animals to the toxic effects of a substance. However, exposure-response data for humans often are not available, and extrapolation from animal data is necessary. The most useful animal studies are those in which the exposure occurs by relevant routes of administration (oral for drinking-water contaminants; dermal absorption and inhalation for contaminants found in water used for purposes of hygiene) and in which the duration of exposure approximates human exposure times. Confidence in extrapolation from animals to humans is increased if at least two non-human species have been examined and if the physiologic disposition (including metabolic pathway), target organs, and toxic effects in animals parallel the effects expected in humans based on available information. NASA will use recently developed techniques for combining uncertainty factors to increase confidence in extrapolations from animal data (Chapter 4). For 1- and 10-day (d) SWEGs, data from acute toxicity tests in animals should be used; for 100- and 1000-d SWEGs, the reference should be subchronic, chronic, or lifetime studies in animals.
When considering the use of published studies on the toxicity of a chemical in establishing a SWEG, attention must be paid to species
and, in some cases, to strain differences as they relate to the applicability of assessing health risks in humans. There might be quantitative or qualitative differences in xenobiotic-metabolizing activity between laboratory animals and humans. There also could be mechanistic considerations to suggest that one species is more predictive than is another of responses in humans.
Exposure via oral feeding or gavage can provide a scientifically sound and defensible basis for estimating effects in humans and for predicting the concentrations at which those effects occur. Studies that use water as the vehicle would be most useful. Repeated exposures to a test substance are useful in the identification of homeostatic adaptations or repair and recovery that could occur over time. It is important to note the doses and to give special attention to evaluation of extremes in dosage. Dosage regimen – gavage versus feeding or drinking water, repeated versus single – and the possibility of additional exposure from inhalation should be considered. Oral dosing experiments permit the testing of hypotheses about the mechanism of the toxic action of the pollutant. Ideally, sufficient data should be collected to establish a no-observed-adverse-effect level (NOAEL) or to establish an accurate estimate of a benchmark dose (BMD). The BMD has a specified low level of excess health risk, generally in the range of 1%-10% that can be estimated from data with little or no extrapolation outside the experimental dose range.
In many studies the substance is administered by gavage on an acute, subchronic, or chronic basis in a vehicle, often water or an oil, in the form of a bolus dose directly into the stomach within a brief period. A major difficulty in evaluating gavage studies is that blood concentrations and attendant effects are induced that might not be observed if the administration were spread out over several smaller doses, as would be expected with the normal pattern for water consumption. The metabolism and pharmacokinetics associated with the single, high doses might be different from what would be observed if repeated, lower doses were used. The resulting absorption might be influenced by the vehicle, or the vehicle itself could have an adverse effect on the animal. The relative absorption from water and food must be considered in evaluating animal studies and in estimating human exposure to contaminants in drinking water and water-reconstituted food.
Subchronic studies that assess regular treatment with the chemical over 90 days (typically) permit examination of cumulative effects while
also permitting time for repair mechanisms or physiologic adaptation. Toxic effects on specific organs can be evaluated as a function of dose and time of administration in a given species, the dose at the target organ, and the likelihood of accumulation of effects over time. Ideally, toxic effects should be related to several factors, including total dose, number of treatments with the test substance, and frequency of administration.
Toxic effects for a specific substance might be different in animals exposed to the contaminant by repeated exposure to low doses over an extended period than they will be in animals exposed to a single, high dose. For example, acute exposure to benzene can cause depression of the central nervous system, bu repeated exposure can result in leukemia. Thus, the most sensitive end point and data for establishing a 1000-d SWEG can be completely different from sensitive end point and data appropriate for establishing a 1-d SWEG.
Chronic studies cover most of the lifetime of an animal, and can range from a relatively short duration to lifetime exposure. They allow examination of effects at low doses and can be used to detect accumulated toxic effects or repair mechanisms that activate after an extended period of exposure. Carcinogenic and noncarcinogenic end points alike can be evaluated. Special attention must be directed to how well results can be extrapolated to humans. Although most effects observed in animals are also seen in humans, there are some examples that do not extrapolate to humans, such as the nephropathy-related appearance of α2u-globulin in rat kidney attendant to exposure to a series of compounds (many of which are components of motor fuels). Another example is the frequent occurrence of primary liver tumor in C3H mice observed in many bioassays despite the fact that primary liver tumor is rarely observed in humans and for the most part is not known to be caused by similar exposures. Data from experiments in which the maximum tolerated dose is used must be interpreted with care because human exposures do not mimic that paradigm of carcinogenesis bioassays. In the interpretation of chronic-exposure bioassays, mechanisms of carcinogenesis, species specificity, examination of threshold levels, and the operation of initiation versus promotion must be integrated to provide the best judgment regarding potential human toxicity.
The same considerations and concerns about subchronic studies apply to chronic studies. Additional attention should be directed to the
influence of naturally occurring conditions in test animals in altering the disposition and the metabolism of the substance during the course of the study. Potential physiologic alterations with age must be addressed.
IN VITRO TOXICITY STUDIES
Important information can be obtained from studies that investigate adverse effects of chemicals on cellular or subcellular systems in vitro. Systems in which toxicity data have been collected include isolated organ systems (e.g., isolated perfused livers and lungs), single-cell organisms, cells isolated from specific organs of multicellular organisms and maintained under defined conditions (e.g., isolated hepatocytes and bone-marrow colony-forming units), functional units derived from whole cells (e.g., organized subcellular particles), breakdown products of cellular disruption (e.g., microsomes and submitochondrial particles), isolated or reconstituted enzyme systems, and specific macromolecules (e.g., proteins and nucleic acids).
In vitro studies can be used to elucidate the toxic effects of chemicals and to provide information on their mechanism of action. Typically, in vitro systems are used on the assumption that effects observed are reasonable models for humans. However, for NASA's purposes, the additional caveat that they should reflect the response of humans in space must be added. Thus, in vitro studies must be interpreted with caution, because the effects of microgravity on cellular systems must be considered.
ADVANCES IN HEALTH EFFECTS ASSESSMENT
In setting SWEGs, all observed toxic effects should be considered: mortality, morbidity, functional impairment, reproductive toxicity, developmental toxicity, genotoxicity, carcinogenicity, neurotoxicity, immunotoxicity, hepatotoxicity, respiratory toxicity, and in vitro toxicity. Developmental toxicity data are included in the analysis for comprehensiveness, even though pregnant astronauts are barred from spaceflight. The various toxicity effects are reviewed extensively in the SMAC guidelines (NRC 1992) and are not repeated here. However,
there have been several advances that are applicable to SWEGs in the areas of neurobehavioral toxicology, reproductive toxicology, and mutagenesis. Those advances are discussed below.
Several reports emphasize the importance of neurobehavioral testing in assessing the effects of chemical pollutants, including water contaminants, on the central nervous system (Moser and MacPhail 1992; MacPhail and Tilson 1995). For the most part, such assessment provides noninvasive measures of sensory-motor performance (speed, accuracy, fine discrimination). Some water contaminants pose a potential health hazard because of their ability to alter nervous system function and impair performance of complex tasks, as shown in studies on occupational exposure (Anger 1990).
Extrapolation from industrial thresholds and other related exposure standards for such determinations is limited by the use of tests based on gross toxic effects under conditions of discontinuous exposure that seldom consider the more subtle effects revealed by performance impairment. Moreover, data on the effects of water contaminants on human performance are rarely available, and the interactive influences of stressful, physically demanding spaceflight environments are undetermined. There is a range of human factors and conditions that can be expected to hamper neurobehavioral adaptation and enhance vulnerability to the toxic effects of environmental pollutants (NRC 1987). The factors include confined space, lack of privacy, weightlessness requirements for readjustment of motor and perceptual skills, disorientation, and space sickness. They occur under conditions of isolation, demanding workloads, and the ever-present danger of being away from Earth.
There is now broad acceptance that functional observational batteries and motor activity tests in animals provide a reliable screen for neurotoxic compounds. These assessments include evaluating clinical signs and measuring motor activity, schedule-controlled behavior, and morphologic change in the nervous system (Sette 1989; Holson et al. 1990; Sette and MacPhail 1992; Moser et al. 1995). Motor activity patterns measured automatically provide a continuous, noninvasive assessment of a water contaminant's effects on a stable performance baseline over an extended interval. In addition, schedule-controlled
behavior based on the programming of performance antecedents and consequences can provide specific measures of learning and memory function (McMillan and Leander 1976; MacPhail 1994) as well as sensory thresholds and reaction times (Brady et al. 1979).
Of particular relevance for detecting and evaluating water contaminants is the increasing recognition of taste aversion conditioning (Garcia et al. 1961; Domjan 1980) as a relatively simple, sensitive, and reliable measure of neurobehavioral effects. When animals are exposed to water contaminated with toxins after they have consumed water flavored with a novel substance (saccharin), they subsequently avoid consuming saccharin-flavored water. Such conditioned flavor aversions have been convincingly demonstrated with a variety of toxic water contaminants, including methylmercury (Levine 1978), cadmium (MacPhail 1982a), cobalt (Wellman et al. 1984), trialkyltins (Leander and Gau 1980; MacPhail 1982b), thallium (Nachman and Hartley 1975; Peele et al. 1986), copper (Nachman and Hartley 1975), arsenic (Rzoska 1953), and lead (Dantzer 1980; Leander and Gau 1980). The results of these studies also suggest that taste aversion procedures might provide a novel and selective paradigm for assessing the interactive effects of water contaminants and purification agents.
The effectiveness of animal tests for predicting the human response to metal chelators, for example, has been demonstrated in reported studies with BAL (British antilewisite or 2,3-dimercaptopropanol) and DMSA (2,3-dimercaptosuccinic acid), a water-soluble BAL analogue (Peele et al. 1987). The finding that BAL was significantly more potent than DMSA in conditioning flavor aversions reflects both the reported differences in the toxicity of the two chelators (Aposhian 1983) and the high and low incidence of clinical side effects (fever, nausea, hypertension) with BAL and DMSA, respectively (Klaassen 1980). There is also a good correspondence between the human response to abusable drugs and preclinical animal tests results with such substances (Brady 1991). Despite the predictive promise of laboratory assessment models, however, more study is needed to evaluate the functional relationship between the results obtained with animals in neurobehavioral toxicity tests and effects in humans.
Morphologic changes in the nervous system can be expected to reveal more serious neurobiologic effects of exposure to environmental contaminants, including water pollutants (e.g., Katz et al. 1981; Johnson and Richardson 1983; Spencer and Schaumburg 1999). Such inva-
sive assessments, however, require both in situ perfusion and labor-intensive contemporary tissue preparation for microscopic examination.
The interpretation and application of reproductive toxicity data are difficult if the intent is to apply results to the determination of SWEGs. First, there are two sexes to consider, and each has a different set of reproductive targets for potential toxicants. Targets include the various cells involved in oogenesis and spermatogenesis, the viability and functional capacity of the mature ovum and sperm, fertilization, implantation, and gestation. The interaction of adenohypophyseal-hypothalamic axis and peripheral reproductive organs can be affected as well. Whereas some of the effects can be grossly and readily observed, others are more subtle and might not become apparent until some time after children of the astronaut are born.
Although there are several tests to evaluate reproductive capacity for each sex, exhaustive studies will not have been performed for most compounds of interest in establishing SWEGs. Nevertheless, for the male, data might be available on morphologic and functional parameters of the testis, epididymis, or accessory sex glands; measurements of semen; or measurements of the concentrations of various hormones. For females, a variety of measurements might be available on effects on morphology or function; the oviduct; or hormonal activity, including effects on the hypothalamus, pituitary, or hormones, such as gonadotropin, chorionic gonadotropins, estrogen, or progesterone. Other measures might include implantation, teratological changes, feto-toxicity, and postnatal determination of impairments. The U.S. Environmental Protection Agency has established guidelines for reproductive toxicity assessment (EPA 1996), and there are several publications that provide guidance on evaluating and interpreting reproductive toxicity end points for human health risk assessment (e.g., ILSI 1999; NRC in press).
It will be necessary to evaluate the dose and regimen of exposure, the route of administration, and the relevance of the test result to the human condition for each toxic effect related to exposure before data can be applied to the calculation of a SWEG.
Available mutagenesis data on a variety of cell types exposed to water contaminants in vivo or in vitro provide information that is useful for determining whether genetic risks should be included in setting concentration guidelines on contaminants in space-station drinking water. Three main types of mutations serve as good indices for mutagenic potential: single-gene mutation, resulting from a change in the molecular structure of DNA, which can result from substitution or loss or gain of a single base pair (genetic polymorphism); loss or rearrangement of large segments of DNA caused by strand breakage (chromosome aberrations); and a change in the amount of DNA in a cell (aneuploidy). In the first two types, DNA is the target molecule; for aneuploidy induction, the proteins in spindle microtubules and centromeres are the most likely targets.
Mutations in germ cells cause heritable abnormalities in offspring. Because these cells in mammals of both sexes are sequestered from circulating blood (more so in the female than in the male), the probability seems small that low concentrations of an ingested chemical will induce mutations in germ cells. In contrast, somatic cells have ready access to blood, as evidenced by the highly mutagenic compounds used for tumor chemotherapy. There are now data showing that a mutation produced in a single somatic cell could result in cancer. Thus, in the space-station environment, mutation that leads to carcinogenesis is the primary concern.
Evidence that mutation is integral to all aspects of carcinogenesis, from initiation through promotion and progression, is now overwhelming (Loeb 1996; Bishop 1997); in fact, mutation is the driving force in the loss of control over cell division and movement. Most tumors are clones; they are derived from mutation in a single cell (Barrett 1995). However, many subsequent mutations, including those that involve mutator genes (Minnick and Kunkel 1996), participate in the complex, multistep, long-term carcinogenic process that leads to a clinically recognizable tumor.
The altered gene products primarily involved in carcinogenesis are generally the proteins that control cell division, growth, and movement, thereby promoting clonal expansion and metastasis. Some of them are transcription factors, tumor suppressors, mitotic check point controllers, chromosome structure stabilizers, DNA damage recogni-
tion proteins and repair enzymes, apoptotic inhibitors, chaperonins, and proteases that break down intercellular adhesion proteins. An example is a mutated tumor suppressor gene, p53, found in 40-50% of common tumors, such as those of the lung, colon, and breast (Perera 1997). Its normal function is to regulate other genes so that growth of a cell with damaged DNA is inhibited until repair is completed or, in the absence of repair, signals production of proteases that kill the cell (apoptosis).
New test systems for mutagenic and carcinogenic chemicals have been developed that considerably improve sensitivity; some of the more prominent are listed here:
The comet assay is a single-cell gel electrophoresis technique for detecting DNA damage (Plappert et al. 1997).
Immunomarkers and multifluorescent probes for specific genes and for all 24 human chromosomes can reveal small increments in mutation frequency. For example, aneuploidy for individual chromosomes can be readily detected in sperm of mammals, including humans (Robbins et al. 1997), and translocation chromosomes can be unequivocally identified in dividing cells (Yang et al. 1997).
Transgenic animals heterozygous for a proto-oncogene (+/−), such as the p53 gene, permit unequivocal in vivo identification of carcinogenic chemicals that cause cancer by mutating one p53 gene; nonmutagenic, noncarcinogenic chemicals also can be identified (Tennant et al. 1996).
DNA site-specific techniques use oligodeoxynucleotides that contain a single chemical-derived DNA adduct that helps identify postsynthetic changes, such as deletions and base substitutions in a gene (Shibutani and Grollman 1997).
The high specificity of polymorphic tumor mutants, and the sensitivity of new tests for identifying them, emphasize the relevance of mutagenesis to chemical carcinogenesis. Detection of low levels of mutagenicity combined with positive biomarker data (Perera 1997) provides strong evidence for mutagenic potency. Data obtained with new methods combined with results from earlier studies enhance confidence in using mutagenicity data in determining SWEGs.
Much attention and research are being devoted to establishing car-
cinogenic risks on the basis of mutagenic data (see Dellarco and Jacobson-Kram 1996; Gold and Zeiger 1996), and there is some optimism that a formulation can be constructed for the recognized vital relationship of the two processes. In the meantime, establishing human risk levels attending germ and somatic cell mutations induced by exposure to various contaminants must rely on the “weight of evidence” method.
Health effect studies are required by regulatory agencies before many substances can be marketed for various purposes in the United States. They fall into the general category of descriptive toxicology. With the maturation of toxicology as a distinct discipline, the investigation of the biology that leads to toxic events has emerged as a significant focus of research. This area of research is called mechanistic toxicology. There is a major effort to apply the results of mechanistic studies to the risk-assessment process. Experiments in mechanistic toxicology might include, but not be limited to, studies of routes of administration, absorption, metabolism, excretion, tissue distribution, formation of biologic reactive intermediates, covalent binding to specific macromolecules, physiologically based pharmacokinetic models, and polymorphisms.
The recent development of investigative tools in molecular biology has opened new ways to evaluate mechanisms of toxicity. It is now possible to study directly the interaction of specific substances with the genome itself, or to study enzymes that modify gene function. The ability to identify and measure a large new class of small proteins, such as the cytokines and related controlling factors, through the use of new methods in immunochemical analysis permits the study of mechanisms by which the toxicity of some substances might be expressed as interfering with the normal functioning of these products. The use of knockout and transgenic animals requires fine judgments to ensure that the data can be extrapolated to normal animals and exposed humans. An excellent example of this application comes from a report on a study of mice in which the gene for the synthesis of CYP2E1 was knocked out. The animals were no longer sensitive to benzene-induced bone-marrow depression, because CYP2E1 is the enzyme re-
sponsible for metabolic activation of benzene to metabolites that inhibit bone-marrow function (Valentine et al. 1996).
Classic toxicity studies use normal animals on Earth. In the controlled conditions of the laboratory, attention is paid to their diurnal rhythms and to light and dark cycles, which might not mimic the situation of NASA's chief concern, the astronaut in space. It might be necessary to conduct laboratory studies under conditions that simulate space flight or to develop animal models with features similar to the physiologic state of the astronaut in prolonged spaceflight. Although such models are approximations to the human condition, they might provide more relevant information than could studies on unaltered normal animals. Animals that have flown in space are likely to be more appropriate surrogates for humans. For example, rats flown aboard Cosmos 1887 showed altered hepatic function (Merrill et al. 1990). They also demonstrated skeletal muscle weakness resulting from muscle fiber atrophy and segmental necrosis. In addition to the microgravity and radiation of space, animals are exposed to launch and reentry gravity forces, noise, and vibration.
Although data from oral exposures to toxic substances are preferred to develop SWEGs for drinking water, data from experiments in which toxicants are administered by other routes are potentially useful. Because species can differ in their responses to toxic substances, the utility of animal data depends in part on the species used. For example, aflatoxin B1 induces liver tumors in rats, hamsters, and monkeys but not in mice (IARC 1993). The mechanistic basis for the difference is thought to be related to species differences in the expression of a particular form of the enzyme glutathione S-transferase in the liver (Eaton and Gallagher 1994). In the absence of information on target organs and pharmacokinetics in both animals or humans, however, the confidence in an extrapolation from animals to humans can be low. As relevant human data are accumulated they should be incorporated into the SWEGs process.
In developing SWEGs, several types of data should be evaluated, including the physical and chemical characteristics of the contaminant, human clinical and epidemiologic studies, in vitro toxicity studies,
toxicokinetic studies, animal toxicity studies conducted over a range of exposure durations, genotoxicity studies, and carcinogenicity bio-assays. All observed toxic effects should be considered, including mortality, morbidity, functional impairment, neurotoxicity, immunotoxicity, reproductive toxicity, developmental toxicity, genotoxicity, and carcinogenicity. For completeness, developmental effects should be considered in the analyses, even though pregnant astronauts are barred from space flight.
Data from oral exposure studies should be used, particularly drinking water and feed studies, in which the duration of exposure approximates human exposure times. Dermal absorption and inhalation studies should also be evaluated, as exposure from those routes can also occur from water.
There are several important determinants for deriving a SWEG, including identifying the most sensitive target organ or body system affected; the nature of the effect on the target tissue; dose-response relationships for the target tissue; the rate of recovery; the nature and severity of the injury; cumulative effects; toxicokinetic data; interactions with other chemicals; and the effects of microgravity.
Anger, W.K. 1990. Worksight behavioral research. Results, sensitive methods, test batteries and the transition from laboratory data to human health. Neurotoxicology 11(4):627-717.
Aposhian, H.V. 1983. DMSA and DMPS – Water soluble antidotes for heavy metal poisoning. Ann. Rev. Pharmacol. Toxicol. 23:193-215.
Barrett, J.C. 1995. Role of mutagenesis and mitogenesis in carcinogenesis. Pp. 21-32 in Environmental Mutagenesis, D.H. Phillips and S. Venitt, eds. Oxford: BIOS Scientific Publishers.
Bishop, J.M. 1997. Cancer: What should be done? Science 278:995>.
Brady, J.V. 1991. Animal models for assessing drugs of abuse. Neurosci. Bio-behav. Rev. 15(1):35-43.
Brady, J.V., L.D. Bradford, and R.D. Heinz. 1979. Behavioral assessment of risk-taking and psychophysical functions in the baboon. Neurobehav. Toxicol. 1(Suppl 1):73-84.
Checkoway, H., N. Pearce, and D.J. Crawford-Brown. 1989. Research Methods in Occupational Epidemiology. New York: Oxford University Press. 344 pp.
Dantzer, R. 1980. Conditioned taste aversion as an index of lead toxicity. Pharmacol. Biochem. Behav. 13(1):133-135.
Dellarco, V.L., and D. Jacobson-Kram. 1996. New Approaches for Cancer Identification and Risk Assessment [Editorial Introduction]. Mutat. Res. 365:1-2.
Domjan, M. 1980. Ingestional aversion learning: Unique and general processes. Pp. 275-336 in Advances in the Study of Behavior, Vol. 11, J.S. Rosenblatt, R.A. Hinde, C. Beer, and M. Busnel, eds. New York: Academic Press.
Eaton, D.L., and E.P. Gallagher. 1994. Mechanisms of aflatoxin carcinogenesis. Annu. Rev. Pharmacol. Toxicol. 34:135-172.
EPA (U.S. Environmental Protection Agency). 1996. Guidelines for Reproductive Toxicity Risk Assessment. Fed. Regist. 61(212):56273-56322.
Garcia, J., D.J. Kimeldorf, and E.L. Hunt. 1961. The use of ionizing radiation as a motivating stimulus. Psychol. Rev. 68(6):383-385.
Gold, L.S., and E. Zeiger. 1996. Handbook of Carcinogenic Potency and Genotoxicity Databases. Boca Raton, FL: CRC Press.
Holson, R.R., M.G. Paule, and F.M. Scalzo. 1990. Methods in behavioral toxicology and teratology. Neurotoxicol. Teratol. 12(5):413-567.
IARC (International Agency for Research on Cancer). 1993. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Some Naturally Occurring Substances: Food Items and Constituents, Heterocyclic Aromatic Amines and Mycotoxins, Vol. 56. Lyon, France: International Agency for Research on Cancer.
ILSI (International Life Sciences Institute). 1999. An Evaluation and Interpretation of Reproductive Endpoints for Human Health Risk Assessment, G. Daston, and C. Kimmel, eds. Washington, DC: ILSI Press.
Johnson, M.K., and R.J. Richardson. 1983. Biochemical endpoints: Neurotoxic esterase assay. Neurotoxicology 4(2):311-320.
Katz, R., C.N. Tai, R.M. Diener, R.F. McConnell, and D.E. Semonick. 1981. Dichloroacetate, sodium: 3-month oral toxicity studies in rats and dogs. Toxicol. Appl. Pharmacol. 57(2):273-287.
Klaassen, C.D. 1980. Heavy metals and heavy-metal antagonists. Pp. 1615-1637 in The Pharmacological Basis of Therapeutics, 6th Ed., A.G. Gilman, L.S. Goodman, and A. Gilman, eds. New York: Macmillan.
Leander, J.D., and B.A. Gau. 1980. Flavor aversions rapidly produced by inorganic lead and triethyl tin. Neurotoxicology 1:635-642.
Levine, T.E. 1978. Conditioned aversion following ingestion of methylmercury in rats and mice. Behav. Biol. 22(4):489-496.
Loeb, L.A. 1996. Many mutations in cancers. Pp. 329-342 in Genetic Instability in Cancer, Cancer Surveys, Vol. 28, T. Lindahl, ed. Plainview, NY: Cold Spring Harbor Laboratory Press.
MacPhail, R.C. 1982a. Flavor aversions induced by cadmium: Importance of route of exposure [abstract]. Toxicologist 2(1):4.
MacPhail, R.C. 1982b. Studies on the flavor aversions induced by trialkyltin compounds. Neurobehav. Toxicol. Teratol. 4(2):225-230.
MacPhail, R.C. 1994. Behavioral analysis in neurotoxicology. Pp. 7-18 in Neurobehavioral Toxicity: Analysis and Interpretation, B. Weiss, and J.L. O'Donoghue, eds. New York: Raven Press.
MacPhail, R.C., and H.A. Tilson. 1995. Behavioral screening tests: Past, present, and future. Pp. 231-238 in Neurotoxicology: Approaches and Methods, L. Chang, and W. Slikker, Jr., eds. San Diego: Academic Press.
McMillan, D.E., and J.D. Leander. 1976. Effects of drugs on schedule-controlled behavior. Pp. 85-139 in Behavioral Pharmacology, S.D. Glick, and J. Goldfarb, eds. St. Louis, MO: C.V. Mosby.
Merrill, A.H.,Jr., M. Hoel, E. Wang, R.E. Mullins, J.L. Hargrove, D.P. Jones, and I.A. Popova. 1990. Altered carbohydrate, lipid, and xenobiotic metabolism by liver from rats flown on Cosmos 1887. FASEB J. 4(1):95-100.
Minnick, D.T., and T.A. Kunkel. 1996. DNA synthesis errors, mutators and cancer. Pp. 3-20 in Genetic Instability in Cancer, Cancer Surveys, Vol. 28, T. Lindahl, ed. Plainview, NY: Cold Spring Harbor Laboratory Press.
Moser, V.C., and R.C. MacPhail. 1992. International validation of a neurobehavioral screening battery: The IPCS/WHO collaborative study. Toxicol. Lett. 64/65:217-223.
Moser, V.C., B.M. Cheek, and R.C. MacPhail. 1995. A multidisciplinary approach to toxicological screening: III. Neurobehavioral toxicity. J. Toxicol. Environ. Health 45(2):173-210..
Nachman, M., and P.L. Hartley. 1975. Role of illness in producing learned taste aversions in rats: A comparison of several rodenticides. J. Comp. Physiol. Psychol. 89(9):1010-1018.
NRC (National Research Council). 1987. A Strategy for Space Biology and Medical Science for the 1980s and 1990s. Washington, DC: National Academy Press.
NRC (National Research Council). 1992. Guidelines for Developing Spacecraft Maximum Allowable Concentrations for Space Station Contaminants. Washington, DC: National Academy Press.
NRC (National Research Council). In press. Evaluating Chemical and Other Agent Exposures for Reproductive and Developmental Toxicity. Washington, DC: National Academy Press.
Peele, D.B., R.C. MacPhail, and J.D. Farmer. 1986. Flavor aversions induced by thallium sulfate: Importance of route of administration. Neurobehav. Toxicol. Teratol. 8(3):273-277.
Peele, D.B., J.D. Farmer, and R.C. MacPhail. 1987. Conditioned flavor aversions: Applications in assessing the efficacy of chelators in the treatment of
heavy-metal intoxication. Toxicol. Appl. Pharmacol. 88(3):397-410..
Percy, C., E. Stanek, 3rd, and L. Gloeckler. 1981. Accuracy of cancer death certificates and its effect on cancer mortality statistics. Am. J. Public Health 71(3):242-250.
Perera, F.P. 1997. Environment and cancer: Who are susceptible? Science 278(5340):1068-1073.
Plappert, U.G., B. Stocker, H. Fender, and T.M. Fliedner. 1997. Changes in the repair capacity of blood cells as a biomarker for chronic low-dose exposure to ionizing radiation. Environ. Mol. Mutagen. 30(2):153-160.
Rinsky, R.A. 1989. Benzene and leukemia: an epidemiological risk assessment. Environ. Health Perspect. 82:189-191.
Robbins, W.A., M.F. Vine, K.Y. Truong , and R.B. Everson. 1997. Use of fluorescence in situ hybridization (FISH) to assess effects of smoking, caffeine, and alcohol on aneuploidy load in sperm of healthy men. Environ. Mol. Mutagen. 30(2):175-183.
Rzoska, J. 1953. Bait shyness: A study in rat behavior. Br. J. Anim. Behav. 1(4):128-135.
Sette, W.F. 1989. Adoption of new guidelines and data requirements for more extensive neurotoxicity testing under FIFRA. Toxicol. Ind. Health 5(2):181-194.
Sette, W.F., and R.C. MacPhail. 1992. Qualitative and quantitative issues in assessment of neurotoxic effects . Pp. 345-361 in Neurotoxicology, H. Tilson, and C. Mitchell, eds. New York: Raven Press.
Shibutani, S., and A.P. Grollman. 1997. Molecular mechanisms of mutagenesis by aromatic amines and amides . Mutat. Res. 376(1-2):71-78.
Smith, T.J. 1987. Exposure assessment for occupational epidemiology.Am. J. Ind. Med. 12(3):249-268.
Spencer, P.S., and H.H. Schaumburg. 1999. Experimental and Clinical Neurotoxicology, 2nd Ed. New York: Oxford University Press.
Tennant, R.W., J. Spalding, and J.E. French. 1996. Evaluation of transgenic mouse bioassays for identifying carcinogens and noncarcinogens. Mutat. Res. 365(1-3):119-127.
Valentine, J.L., S.S. Lee, M.J. Seaton, B. Asgharian, G. Farris, J.C. Corton, F.J. Gonzalez, and M.A. Medinsky. 1996. Reduction of benzene metabolism and toxicity in mice that lack CYP2E1 expression. Toxicol. Appl. Pharmacol. 141(1):205-213.
Wellman, P.J., P.A. Watkins, J.R. Nation, and D.E. Clark. 1984. Conditioned taste aversion in the adult rat induced by dietary ingestion of cadmium or cobalt. Neurotoxicology. 5(2):81-90.
Yang, T.C., K. George, A.S. Johnson, M. Durante, and B.S. Fedorenko. 1997. Biodosimetry results from space flight Mir-18. Radiat. Res. 148(Suppl. 5):S17-S23.