Guidelines for Developing Spacecraft Maximum Allowable Concentrations for Space Station Contaminants (1992)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Establishment of Spacecraft Maximum Allowable Concentrations DESCRIPTION OF SMACs Spacecraft maximum allowable concentrations are used as guidance for chemical exposure, either during normal operations of the space- craft or during emergency situations. A 1- or 24-hr SMAC is defined as a concentration of a substance in air (such as a gas, vapor, or aerosol) that may be acceptable for the performance of specific tasks during emergency conditions lasting for periods of less than 1 hr or less than 24 hr. The effects of an exposure at a 1- or 24-hr SMAC could include reversible effects that do not impair judgment and do not interfere with proper responses to the emergency. The kinds of emergency exposures anticipated could result from events such as fires, spills, or line breaks. The 1- and 24-hr SMACs are acceptable only in an emergency when some risks or some discomfort must be endured to prevent greater risks (such as fire, explosion, or massive release). Even in an emergency, exposure should be limited to a defined short period. Exposure at the 1- and 24-hr SMACs might produce such effects as increased respiratory rate from increased carbon dioxide, headache or mild central nervous system effects from carbon monoxide, and res- piratory tract or eye irritation from ammonia or sulfur dioxide. The 1- and 24-hr SMACs are exposure levels that should not cause serious or permanent effects. While minor reduction in performance is per- missible, it should not be so much as to prevent proper responses to the emergency (such as shutting off a valve, closing a hatch, removing a source of heat or ignition, or using a fire extinguisher). For exam- ple, in normal work conditions, a degree of upper respiratory tract irritation or eye irritation causing discomfort would not be considered acceptable; during an emergency, such an effect would be acceptable if it did not cause irreversible harm or seriously affect judgment or performance. 59

60 GUIDELINES FOR DEVELOPING SMACS SMACs for up to 180 days are concentrations designed to avoid adverse health effects, either immediate or delayed, and to avoid deg- radation in performance of crew after continuous exposure. In con- trast to 1- or 24-hr SMACs, which are intended to guide exposures during emergencies (exposures that, although not acceptable under normal operating conditions, should not cause serious or permanent effects), 180-day SMACs are intended to provide guidance for opera- tions lasting up to 180 days in an environment like that of a space station. Accumulation, detoxification, excretion, and repair are im- portant in determining 180-day SMACs. If a material is cumulative in its effects, its 180-day SMAC must take that into account. Neuro- pathological regeneration or repair of toxic injuries occurs more readi- ly in intermittent exposures than in continuous exposure to constant toxic insult. Therefore, repair is important in the recommendation of 180-day SMACs. SOURCES OF DATA FOR DEVELOPING SMACs Various types of evidence should be assessed in establishing SMAC values. These include information from (1) chemical-physical charac- terizations of the potential toxicant, (2) in vitro toxicity studies, (3) animal toxicity studies, (4) human clinical studies, and (5) epidemic- logical studies. Chemical-Physical Characteristics of Toxicant The chemical and physical characteristics of a chemical provide valuable information on the dosimetry of the compound within the body and on the likely toxic effects. For example, size and water solubility of inhaled particles strongly influence where the material deposits in the respiratory tract. Likewise, lipophilicity influences absorption of the material where it accumulates and how long it re- mains in the body. Structure-activity relationships may allow esti- mation of the toxic potential of new compounds based on the known toxicities of well-investigated structurally related compounds. How- ever, additional uncertainty (safety) factors must be applied to arrive at safe levels for those congeners that have no dose-response data from intact animals.

ESTABLISHMENT OF SMACS 61 In Vitro Toxicity Studies Important data 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 including bacteria, cells isolated from specific organs of multicellular organisms and maintained under defined conditions (e.g., isolated hepatocytes and bone-marrow colony-forming units), func- tional units derived from whole cells (e.g., organized subcellular par- ticles including nuclei and mitochondria), breakdown products of cel- lular disruption (e.g., microsomes and submitochondrial particles), isolated or reconstituted enzyme systems, and specific macromolecules (e.g., proteins and nucleic acids). Data on inhibition of specific physiological functions, pathological outcomes of exposure, genetic damage, changes in xenobiotic metabo- lism, or changes in levels or quality of cellular components can con- tribute to these evaluations. In vitro studies can help both to describe the effects of chemicals and to provide information on the mechanism of action of chemicals. In vitro systems are used on the assumption that the effects ob- served present a reasonable model for humans. In the current context, the additional caveat that they should reflect the response of humans in space must be added. Therefore, use of older data and plans to collect new data should pay heed to the need for surrogate modeling relevant to the astronaut. Animal Toxicity Studies The data necessary to evaluate the relationship between exposure to a pollutant and its effects on a population are frequently not available from human experience. For many air pollutants, studies in animals have provided the only useful data. Ideally, the data should be de- rived from at least two species and by the inhalation route. Inhalation experiments with animals provide a basis for estimating possible effects in humans and the concentrations at which these ef- fects occur. They are useful in the identification of adaptations that may occur following repeated exposure. They permit the testing of hypotheses about the mechanism of the toxic action of pollutants.

62 GUIDELINES FOR DEVELOPING SMACS They offer a good opportunity to explore interactions between pol- lutants and other factors that may affect toxicity. Data from skin absorption, ingestion, and parenteral studies are also potentially useful. Since eye irritation can be debilitating, eye-irritan- cy testing of substances found in space-station air also is needed. The usefulness of animal data depends in part on the species used. Rele- vance to humans may be limited in the absence of information on tar- get organs and pharmacokinetics in both animals and humans. The better animal studies will report the following: • The most sensitive target organ(s) or body system(s) affected by exposure to the contaminant in question. • The nature of the effect on the target organ(s). • Data to establish dose-response relationships for the target organs(s)—from no effect to severe effects. (The distinction between exposure and dose needs to be made.) • The rate of recovery from reversible effects, if any. • The nature and severity of injury for effects that are not revers- ible. • Cumulative effects, if any, such as neurotoxicity and cancer. • Pharmacokinetic data for comparison with data obtained from humans. • The effects of interaction, if any, of the toxicant with other air pollutants (or exposure conditions) and the minimum concentrations at which the interaction appears to occur. • Techniques used to assure quality and avoid bias. Classic toxicity studies employ normal animals. It may be necessary to develop animal models with features similar to the physiological state of the astronaut in prolonged spaceflight. Although such models are approximations to the human condition, they should provide better information than studies on unaltered normal animals. Thus, animals flown in the space shuttle or the space station are likely to be more appropriate surrogates for humans. For example, rats flown aboard Cosmos 1887 showed altered hepatic function. These rats also demon- strated skeletal muscle weakness resulting from muscle fiber atrophy and segmental necrosis. Studies of the myocardium showed evidence of atrophy. In addition to the microgravity of space, animals can also be ex- posed to launch and reentry gravity forces, noise, and vibration.

ESTABLISHMENTOF SMACS 63 Clinical Studies In establishing SMACs for chemicals, dose-response data from human exposure are most desirable, and such data should be used whenever possible. Experimental human studies can be designed to provide useful information on dose-response relationships. Ethical concerns limit these studies to pollutants that are anticipated to have no residual effects consequent to the experimental exposure and to short-term studies. Data from inhalation exposures are most useful here, because inhalation is the most likely route of exposure. Epidemiological Observations Human toxicity data frequently are obtained from epidemiological studies of long-term industrial exposures as well as short-term exposures usually to high levels of toxicants following accidents. These data sometimes provide a basis for estimating a dose-response relationship. Epidemiological studies have contributed to our knowledge of the health effects of many airborne chemical hazards, for example, radon daughters (Whittemore and McMillan, 1983) and vinyl chloride (Waxweiler et al., 1976). Studies of environmental and occupational exposure can assess acute effects of short-term exposure such as myo- cardial infarction after exposure to methylene chloride (Stewart and Hake, 1976) or effects of long-term exposures such as cardiovascular disease associated with long-term exposure to carbon monoxide (Stern et al., 1988). One limitation of most epidemiological studies is the limited information that is available about past exposure (Checkoway et al., 1989). Studies that include estimates of past exposure based on available historical exposure records, either of the work environment or personal samples, tend to be more useful than those based on years of employment (Rinsky, 1989). Studies that involve assessment of exposure that occurs while the cohort is followed are likely to 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 compli- ance with exposure regulations (Smith, 1987). However, an appealing aspect of retrospective studies is that the exposure, the interval be- tween exposure and onset of disease, and the onset of disease will have occurred by the time the study is conducted; in prospective studies,

64 GUIDELINES FOR DEVELOPING SMACS time must pass for any responses to occur. Epidemiological studies also vary in the accuracy and precision of the health outcome meas- ured. Some of these outcomes must rely on available information such as death certificates, which may be incorrect at times (Percy et al., 1981), while others rely on clinical testing, pathological reports, or early preclinical markers of pathology. Despite these limitations, if the populations studied are large enough, had substantial exposure, and had sufficient interval between exposure and study to allow for the expression of disease, epidemiological studies have the major ad- vantage of considering the effects of exposure in humans. Epidemio- logical studies can often provide the basis for establishing a permis- sible concentration for human exposure. Epidemiological outcomes often are reported in terms of relative risk. Relative risk is a ratio of the rate of outcome of either disease or disability in the exposed population to that in the nonexposed pop- ulation. Relative risk is not a measurement of risk. The relative risk for a rare disease can be equivalent to that of a common disease and lead to substantially less total risk. To establish a level of acceptable exposure, information on relative risk usually must be reformulated to understand the relationship of relative risk and risk of morbidity or mortality. The level of risk that is acceptable is a matter of policy, not epidemiology. TYPES OF DATA USED IN RECOMMENDING SMACs The preceding section described the sources of data for use in es- tablishing SMACs. In this section, the two types of data obtained from these sources and their importance in establishing SMACs are described. The types are (1) dosimetry and (2) toxicity end points. Dosimetry Deposition of Particles, Gases, and Vapors in the Respiratory Tract Under Microgravity Conditions Aerosols deposit in the respiratory tract of people primarily by the processes of sedimentation, impaction, interception, and diffusion. Sedimentation is due to the gravitational force acting on the particles,

ESTABLISHMENT OF SMACS 65 and the particles may settle and deposit on the lower surfaces of air- ways. Impaction is due to the inability of individual particles to fol- low the curvature of air streamlines because of inertia, and the par- ticles may hit and stick on the walls of airways. Diffusion is due to random (Brownian) motion of small particles caused by interaction with air molecules, which may cause a particle to move across the air streamlines and deposit on contact with the airway wall. In a micro- gravity environment, sedimentation will not be a mechanism of depo- sition, and deposition will be mainly due to inertial impaction and diffusion. For small particles (<0.2 ^m), deposition is dominated by the dif- fusion mechanism, which is independent of gravity. Gas and vapors also come in contact with airway walls due to Brownian motion. Thus, the deposition of small particles, gases, and vapors in the respiratory tract will not be significantly affected by microgravity conditions. As the particle size increases (1-5 nm range), sedimentation becomes an increasingly important mechanism of deposition under normal gravity forces. Therefore, fractional deposition of particles >1 pm in the respiratory tract will be significantly less under microgravity condi- tions. The pattern of deposition also will shift so that proportionally more of the large particles will deposit in the tracheobronchial region, as opposed to the alveolar region of the respiratory tract. In addition, sedimentation is important for particles >5 nm because it greatly af- fects the persistence of these particles in air. The concentration of particles from 5 to 100 |tm and greater inside the spacecraft will be much higher under conditions of microgravity. These large particles are likely to be very irritating to the eye and the respiratory tract. Transmission of infections can increase because large droplets of saliva and respiratory secretions remain suspended in the atmosphere for a longer time under microgravity conditions than under normal condi- tions. Changes in total deposition, regional deposition, and concentra- tion of large particles in air could influence the potential health ef- fects of particles in the space environment. Thus, the absence of gravity makes a significant difference in par- ticle deposition only when particles range in size from 0.5 to 2 jun. Diffusion will account for most deposition for vapors and for particles smaller than 0.5 urn, and impaction will be responsible for most par- ticle deposition for particles larger than about 2 pm in aerodynamic diameter.

66 GUIDELINES FOR DEVELOPING SMACS Pharmacokinetics and Metabolism The evaluation of the health effects of any chemical in a given environment is often improved by an understanding of its physiologi- cal disposition in the body (pharmacokinetics and metabolism). Metabolism of a chemical may lead to detoxication or may result in metabolic activation leading to toxic effects. Since metabolic events are usually enzyme mediated, they are driven by the concentration of substrate available for the reaction. The concentration of substrate is a result of the level of exposure and the pharmacokinetics of the chemical. Modern pharmacokinetic studies are increasingly aimed at developing dispositional models, which compare input of chemical with outflow for the whole system as well as for individual organs. The space station is a system that can be modeled for any given chem- ical. The space station is a closed system with limited capacity to clear the air of chemical vapors, while the crew contributes to the removal of some chemicals through sequestration and metabolism. Thus, phys- iologically based pharmacokinetic models of the disposition of inhaled materials should be useful in helping to assess the risk of disease from airborne toxicants in the space station. The toxic metabolites of each chemical and factors that control the rate at which they form are important for SMACs. It is important to distinguish between the production of toxic metabolites in organs where toxicity is observed and metabolism in liver where most chemi- cals are metabolized. The relationship between metabolism in liver and toxic effects in either liver or other organs is an important factor because metabolism of chemicals in a liver already damaged by other exposures is likely to differ from metabolism in a normal liver. Fur- thermore, if toxic metabolites are moved from the liver to other or- gans, damaged livers may alter the production of toxic responses (Merrill et al., 1990). To determine the possible biochemical effects of prolonged weight- lessness on liver function, samples of livers from rats that were on board Cosmos 1887 were analyzed for a number of key enzymes in- volved in metabolism of compounds and xenobiotics (Merrill et al., 1990). They observed slightly lower than normal amounts of cyto- chrome P-450 in the livers of the flight group; this finding was similar to that of Spacelab 3 (Merrill et al., 1987). They also observed de- creases in the enzymes aniline hydroxylase and ethylmorphine-Af- demethylase. It is not known with certainty whether spaceflight alters the body's ability to metabolize drugs; however, these limited results

ESTABLISHMENT OF SMACS 67 suggest that a change in the array of cytochrome P-450 enzymes oc- curred in the livers of the animals. This family of enzymes is respon- sible for the metabolism of drugs, the metabolism of a number of natural steroid hormones and their metabolites, and the metabolism of several compounds and mediators critical for intracellular com- munication and regulations, e.g., prostaglandins and various growth factors. An important function of these compounds appears to be their role in the hematopoietic system. Therefore, the normal func- tioning of the cytochrome P-450 system could have important impli- cations for space biomedicine. Toxic metabolites have been observed to be highly reactive chemi- cally and to react with nucleic acids, proteins, or lipids to alter normal biological function. These metabolites may induce alterations in DNA replication or the process of transcription. Attempts to repair damage to DNA may involve misrepair, leading to erroneous DNA replication or function, inhibition of protein synthesis if RNA is the target, or in- hibition of the enzyme or other activity if proteins are the target. In addition to the formation of reactive intermediates that are metabolites of the chemicals, metabolic activity may give rise to species of active oxygen, which may damage nucleic acids or proteins or yield lipid peroxidation. The effects may range from target-organ toxicity to carcinogenesis. Biological Markers Recently, the concept of biological markers as indicators of ex- posure to polluted air has been investigated (NRC, 1989b). Biological markers within an exposed individual can indicate the degree of ex- posure to a pollutant, the initial structural, functional, or biochemical changes induced by exposure, and, eventually, the changes associated with adverse health effects. Biological markers are indicators of change within an organism that link an exposure to polluted air to subsequent development of an ad- verse health effect. It is convenient to divide biological markers into three groups: (1) biological markers of exposure, (2) biological mark- ers (early predictors) of effects of exposure, and (3) biological markers of susceptibility to effects of exposure. Biological markers of exposure can be thought of as "footprints" that the chemical leaves behind after interaction with the body. Such markers contain the chemical itself or a metabolic fragment of the

68 GUIDELINES FOR DEVELOPING SMACS chemical and, thus, are generally chemical-specific. Examples of such markers are the chemical adducts formed between alkylating agents and macromolecules, such as nucleic acids or proteins, particularly blood proteins. Another example is the presence of a volatile chemical in exhaled air. This type of marker has been used by Wallace (1987) to assess exposure to benzene during filling of gasoline tanks in passenger cars, exposure to tetrachloroethylene in dry-cleaning shops, exposure to chloroform from contaminated hot shower water in homes, and exposure to volatile aromatic compounds in tobacco smoke. The breath of crew members of the space shuttle or the space station could be assayed for volatile chemicals at the end of a tour of duty to assess exposures during spacef lights; similar measurements are made on submariners on extended cruises (Knight et al., 1984, 1985). Physiologically based pharmacokinetic models (Ramsey and Andersen, 1984) have been used to relate biological markers of exposure to prior exposure conditions. Biological markers of the effects of inhalation exposure can be any indication of a chemically induced disease. The biological markers of greatest interest are those that are early predictors of late-occurring effects. Such markers would be invaluable in assessing what levels of pollutants can be tolerated in the space station without causing irre- versible deleterious health effects. For example, cell proliferation may result in clonal expansion of initiated cells (Swenberg, 1989). Thus, persistent cell proliferation could be a biological marker predictive of an increased incidence of late-developing neoplastic lesions. Few markers, however, have been validated as predictive of late-stage diseases. The third type of biological marker, markers of increased suscep- tibility to the effects of exposure to airborne chemical pollutants, is a potentially important, useful tool. Such markers possibly could be used to predict which persons are more likely to be adversely affected by space-station exposure. For example, polymorphisms related to acetylation and DNA repair can be related to susceptibility to chemi- cally induced tumors. Such indicators of susceptibility must be used cautiously, however, as they pose numerous moral and ethical prob- lems. The appearance of metabolites in excreta have been useful in the past and will be even more useful in the future as biological markers of exposure, effect, and susceptibility. Metabolites that are specific to the chemical will be most useful. When similar compounds related to normal diet are found, the interpretation of the results may be more

ESTABLISHMENT OF SMACS 69 complex. Other types of biological markers that are undergoing eval- uation include evidence of chromosomal damage and covalent binding of reactive metabolites to DNA or proteins in circulating blood cells. Toxicity End Points: Humans and Animals Mortality Short-term exposure to a pollutant at high concentrations may result in death, which may be immediate or delayed. Mortality is a widely used and important index in animal experiments and is also most use- ful in epidemiological studies as an index of serious health effects. All causes of death may be used as data for SMACs; however, certified causes of death in humans are inaccurate at times, making it more difficult to associate exposures with toxic outcomes. Morbidity (Functional Impairment) Pollutants also produce functional impairment. Most of the con- siderations that apply to mortality as useful data also apply to mor- bidity. It is usually much more difficult to obtain reliable information on morbidity than on mortality. Whereas statistics of mortality are collected routinely on whole populations, information on morbidity usually is not. Health surveys such as the Health Interview Survey (HIS) and Surveillance Epidemiology and End Results (SEER) (cancer data) collect information on medical effects. Often, studies must be initiated to answer the exposure-response questions rather than relying on information collected for other purposes. Clinical Signs and Symptoms Physical examination and paying attention to signs and symptoms can contribute to the biomedical and behavioral assessment of expo- sure. Eye irritation and tearing, for example, are among the most sensitive indicators of excessive exposure to oxidants. Complaints of discomfort by exposed individuals provide an important criterion on acceptability of pollutants. Although they may not be related to any direct health effects, complaints of discomfort need to be considered

70 GUIDELINES FOR DEVELOPING SMACS as an indication of exposure and as a factor in setting acceptable expo- sure levels. Pulmonary Effects The main portal of entry of air pollutants is the respiratory tract. Respiratory symptoms, allergic sensitization, or changes in lung func- tion provide evidence of the impact of pollutants. Spirometry is used by far the most often of the large number of tests used to measure changes in lung function. Widely used measures of the bellows func- tion of the lung include the volume of air that can be expelled after a full inspiration (the forced vital capacity); the volume that can be expelled in a measured time, usually 1 sec (the timed vital capacity or forced expiratory volume in 1 sec, FEVj); the ratio of the forced ex- piratory capacity to the forced vital capacity (expressed as a per- centage); lung compliance; and inspiratory and expiratory resistances. Other useful tests are ventilation rate, gas exchange, and blood flow. A recent report from the NRC (1989b) on pulmonary markers gives detailed information on the use of some of these measures. Hepatic Effects An excellent source for reviewing the adverse effects of chemicals on the liver is by Plaa (1991) in Casarett and Doull's Toxicology. The liver is the single most important organ concerned with protecting the body against the invasion of foreign, potentially toxic chemicals and, therefore, is among the organs most vulnerable to chemical insult. The liver avidly accumulates many chemicals, often after oral admin- istration, leading to the so-called "first pass effect." However, accu- mulation of chemicals and toxic responses of the liver may also occur after inhalation or exposure via other routes. Some chemicals, such as phosphorus, may be directly toxic to the liver. Chemicals passing through the hepatic circulation may be acted upon to yield detoxification products or may be activated metabolical- ly to become more toxic chemical entities. Metabolic activation of many chemicals leads to liver damage. Mechanisms by which liver injury may occur include the accumulation of lipids, lipid peroxida- tion, covalent binding of reactive metabolites to critical cellular mac- romolecules, depletion of antioxidants such as glutathione leading to

ESTABLISHMENT OF SMACS 71 oxidative stress, interactive toxicological effects originating in Kupffer cells that lead to damage in hepatocytes, disruption of cal- cium compartmentation, and other aberrations that may lead to cell death. The effects of chemicals may include local or generalized lesions leading to cell death and necrosis, inflammation, fatty infiltra- tion, cholestasis, cirrhosis, or carcinogenesis. Underlying these injuries may be factors such as excessive alcohol intake, diabetes mellitus, or starvation that induce increases in the levels of cytochrome P-450HE1, an enzyme involved in the metabolic activation of many halogenated hydrocarbons and other chemicals to lexicologically active or carcinogenic species. Some halogenated hydrocarbons, such as halothane, can cause immunological responses in liver following repeated exposure. Nutritional deficiencies may exacerbate potential liver injury caused by chemicals such as ethanol. Liver damage may be detected by the appearance of jaundice, measurement of hepatic enzyme levels in circulating blood, use of radioisotope accumulation in the liver, use of various imaging tech- niques, or, more invasively, evaluation of liver samples obtained by biopsy. Reproductive and Developmental Effects The reproductive and developmental effects of exposure to space- craft toxicants could be either overt or subtle, and the full complement of potential reproductive effects that could occur in the space station is not known. Adverse effects on factors such as reproductive capaci- ty or behavior in offspring of returned astronauts could escape detec- tion easily against the background of human variation. These factors need careful consideration in risk estimation. An additional factor is that components of experimental packages are yet to be designed, although many of the supplies that are likely to be on board are known or will be determined. Each component merits careful consideration, and some additional testing may be needed because reproductive and developmental toxicity data are frequently absent from toxicological data bases (NRC, 1984d). The potential effects of spacecraft toxicants on reproductive out- comes could be diverse both in nature and in mechanism. Examples of potential adverse effects in humans can be derived in large part from animal studies. These effects include testicular atrophy (Melnick, 1984), transient reduction in the number of spermatozoa

72 GUIDELINES FOR DEVELOPING SMACS (Hanley et al., 1984), and in utero developmental problems or post- natal abnormalities of offspring (Kimmel and Buelke-Sam, 1988). For the purposes of this report, two assumptions may be useful concerning reproductive and developmental toxicities: (1) in general, adverse effects on reproduction and in utero development are threshold phenomena, and outside of direct-acting mutagens, exposures below certain levels have not been associated with adverse outcomes; and (2) an excess of any contaminant may cause adverse outcomes. Acute exposure to chemicals at levels that produce overt adult toxicity, defined as disruption of homeostasis (Skalko and Johnson, 1987), may be considered as also capable of interfering with reproduc- tion and development. The assumption could perhaps be made that such severe toxicity scenarios would not occur, but if they did, adverse effects on reproduction also would be a concern. After the exposure had ended and normal homeostasis was established, the reproductive system or a conceptus could have transient or permanent sequelae that could be subtle and delayed in appearance and be unrecognized as attributable to space-station-derived exposures. The agents that merit special attention are those that interfere with reproduction or development at exposure levels too low to be consid- ered acutely toxic. The concept of target-organ toxicity can be used to focus attention on those chemicals whose most vulnerable target is reproductive function or in utero development. Examples of such chemicals are ethylene glycol monomethyl ether, which has one stage of spermatogenesis as its most vulnerable target, and thalidomide, which, although having exceptionally low toxicity for adults, has an as yet undiscovered aspect of in utero development as its most vul- nerable target. Details of test protocols for reproductive and developmental effects testing are readily available from the U.S. Environmental Protection Agency (EPA, 1985). Guidelines for developmental toxicity risk assessment (EPA, 1991) and proposed guidelines for reproductive risk assessment (EPA, 1988a,b) discuss interpretation of data in this area. Neurobehavioral Effects Some pollutants have such subtle effects on the central nervous system (CNS) that they are detectable only by special behavioral tests (McMillan, 1987). For the most part, these tests are designed to meas- ure sensorimotor performance—e.g., speed, accuracy, and fine dis-

ESTABLISHMENT OF SMACS 73 crimination (NRC, 1984d). A large number of prevalent spacecraft contaminants represent a potential major hazard because of their capa- bility to alter CNS function and impair the performance of complex tasks (Anger, 1984). There are compounds for which extrapolating industrial threshold limit values (TLVs) to establish SMACs is inap- propriate particularly if the industrial standards are based on gross toxic effects of the compounds and involve discontinuous exposures. Subtle effects, such as performance impairment, seldomly have been considered in establishing industrial TLVs. Data on the effects of contaminants on human performance are rarely available. There is broad agreement, however, that screening for neurobe- havioral toxicity can be carried out effectively with laboratory animals by measuring motor activity, schedule-controlled behavior, and mor- phological change in the CNS (Holson et al., 1990). The automated measurement of motor-activity patterns provides a continuous nonin- vasive assessment of a pollutant's effects on a stable performance baseline over an extended time interval (Dews, 1953; Reiter, 1977; Reiter and MacPhail, 1979). Schedule-controlled behavior, based on procedures for programming performance antecedents and consequen- ces, can be used to measure specific memory and learning functions as well as sensory thresholds and reaction times (McMillan and Leander, 1976). The measurement of morphological change in the CNS can reveal serious neurobiological toxicity produced by environ- mental contaminants (Spencer and Schaumburg, 1980). Such assess- ments, however, require in situ perfusion and the use of contemporary tissue preparation for examination by light microscopy and, prefer- ably, electron microscopy. Carcinogenicity Since the first manned spaceflights almost 30 years ago, the conse- quences of exposure to potential carcinogens (chemicals and radiation) have been a concern. To date there is no evidence that malignant disease occurs more frequently in astronauts and cosmonauts than in a similar cohort of relatively young persons. However, only a limited number of astronauts have been exposed, and to find significant ex- cess cancer among them would imply a large increase in rates. It is widely accepted that cancer induction is a multiphasic process with at least three broad stages: initiation, promotion, and progression. Clearly, exposure to potent mutagens and known initiators must be

74 GUIDELINES FOR DEVELOPING SMACS avoided in the space-station environment. Identification of cancer promoters is more difficult and their implication in human-cancer induction continues to be a source of controversy. Chemical contam- inants in the air of the space station should be assessed by accepted methods for their tumor-promotion potential. In general, promotion appears to be a dose-dependent phenomenon. However, establishing thresholds for promoters poses a significant problem for lexicologists. The carcinogenic potential of chemical mixtures in the space-station environment needs to be considered. The carcinogenicity of mixtures at low concentrations is frequently assumed to be additive. Mutagenicity Nonlethal mutations induced by chemical or physical agents accu- mulate with prolonged or repeated exposure conditions (long-term effects) such as those expected on the space station. In general, a gene mutation involves a molecular change within a single gene (point mu- tation); a chromosomal mutation involves blocks of genes and results from breakage (altered microscopic structure resulting in a chromosome structural aberration) or from nondisjunction (aneu- ploidy). Somatic-cell mutations may cause damage in the exposed individual (carcinogenesis), whereas germ-cell mutations may cause abnormalities in future generations. Rapidly accumulating data show that mutations in single genes or chromosomes are directly related to cancer initiation, promotion, and progression (Adams and Cory, 1991; Solomon et al., 1991; Weinberg, 1991) and to heritable developmental defects that result in miscarriage or offspring with one or more abnormalities (McKusick, 1990; Borgaonkar, 1991; Rinchik, 1991; Pawson, 1991). This impressive evidence for the impact of single mutations in humans makes impera- tive the inclusion of mutagenic data in assessing the risks of space- station contaminants. Several approaches are used to gather the information needed to determine whether genetic risks should be included in deriving SMACs for given chemicals: 1. Mutagens expected on board. A literature search on the muta- genicity of specific compounds would reveal whether sufficient dose

ESTABLISHMENT OF SMACS 75 (concentration x exposure time)-response data are available for making risk estimates. Data from short-term mutagen assays can be helpful in establishing permissible emergency peak concentrations as well as maximum allowable concentrations during long-term exposure. 2. Human cell studies. The human is a good "filter" for chemicals in air. Three cell types^the peripheral lymphocyte, the erythrocyte, and sperm—are human cells readily accessible for analysis of gene and chromosomal mutations and represent somatic and germ cells. In addition to mutation, molecular biomarkers can be detected in these cell types. Such markers are indicative of damage to crucial mol- ecules. • Lymphocytes (T cells) are excellent for detecting chromosomal changes and have been widely used for in vivo and in vitro mutagen assays to examine structural and numerical aberrations, sister chroma- tid exchanges, micronuclei, unscheduled DNA synthesis, and single- strand breaks in DNA. Few data are available, however, on chromo- somal mutation in lymphocytes of astronauts (Gooch and Berry, 1969; Lockhart, 1974, 1977), and in light of current analytical methods, the data are inconclusive. A fair amount of information exists, however, about the response of lymphocytes to the conditions of spaceflight. Analysis of blood from astronauts on the first 12 shuttle flights revealed a significant decrease in the number of circulating lympho- cytes (NASA, 1989). The ability of lymphocytes to be activated (undergo DNA synthesis and cell division) by mitogens (foreign anti- gens) is greatly diminished also or, in some cases, obliterated, as first reported after the 1961-1969 Soyus-6, -7, and -8 flights (Konstantinova et al., 1973) and later confirmed after the Skylab and Spacelab DI flights (Kimzey, 1977). Although the number and func- tion (activation capability) of lymphocytes returned to normal after the flights, there is concern about the consequences of long-term mis- sions. Experiments suggesting that those effects on lymphocytes are caused by microgravity have shown that exposure of lymphocytes to 1 g in flight (centrifuge) produced no changes when compared with lymphocytes at microgravity (in vivo and in vitro) (NASA, 1989). Hypergravity (10 g) increases the response of lymphocytes to mitogens by as much as 500% over the level observed at 1 g (Lorenzi et al., 1986). Physical stress associated with marathon running causes the

76 GUIDELINES FOR DEVELOPING SMACS same effects on lymphocytes as does the stress of spaceflight (Gmunder et al., 1988). Of special interest with respect to chemical effects are reports that exposure to nitrogen dioxide at concentrations as low as 1.0 ppm suppresses T- and B-lymphocyte responses (Fenters et al., 1973; Maigetter et al., 1976; Richters and Damji, 1988). • Erythrocytes and T lymphocytes are useful for detection of in- duced gene mutations (Albertini et al., 1990). The end-point muta- tions in erythrocytes are mainly in hemoglobin and GPA (glycophorin A) genes and in T lymphocytes, HGPRT (hypoxanthine-guanine phosphoribosyl-transf erase) and HLA (human leukocyte antigen) genes. • Human sperm chromosomes can be examined by fusing them with hamster eggs; this method has been used to study the mutagenic effects of radiation and some chemicals (Brandriff and Gordon, 1990). Detection of gene mutation in sperm is also possible (Wyrobek et al., 1990). Analysis of sperm head shape and motility provides a measure of the damage to male germ cells (Wyrobek and Bruce, 1978). Mutations in either male or female germ cells may lead to reproduc- tive and developmental toxicity (Kay and Mattison, 1985; Goldsmith et al., 1984). If toxicity were to occur, the effects usually would be difficult to detect and quantify in a given person because of variations among individuals (e.g., sperm count), because of time and manner of expression (e.g., congenital defects) and especially because of similar effects occurring spontaneously in other offspring. These difficulties should not diminish concern about their occurrence. The induction of mutation may occur at any stage of spermatogen- esis and result in a mutant sperm appearing immediately or years later. In a female, all but the final stages of oogenesis are completed pre- natally, so that at birth the ovary contains only primary oocytes. Ex- pression of a mutation induced in a primary oocyte could be delayed for years—i.e., delayed until the follicle containing it matures to ovula- tion and the fertilized egg develops into a fetus or even adulthood. • Molecular biomarkers that are primarily carcinogen-DNA or carcinogen-protein adducts provide an extremely sensitive test of cell interaction with chemical contaminants. Minute amounts of such markers can be detected in DNA, e.g., 1 adduct in 109-1010 nucleo- tides; protein adducts in body fluids (blood, urine, breast milk, and semen), exhaled air, and adipose tissue; and new proteins, such as those produced by activated oncogenes (mutated genes that cause can- cer) (Perera et al., 1991).

ESTABLISHMENT OF SMACS 77 Immunotoxicology Immunotoxicology is a science that explores the effects of chemical agents and other harmful substances on the immune system. The immune system can be either stimulated or inhibited by xenobiotics. Excessive stimulation can result in hypersensitivity or autoimmunity, and suppression can be expressed by an increased susceptibility of the host to infectious or neoplastic disease. Xenobiotic-induced immune dysfunction has been well established for several chemicals in animals. In some cases, the immune system has been identified as the most sensitive target organ to detect the minimum toxic doses of a xenobi- otic. Although one or more of the many compartments of the immune system may be significantly suppressed, this suppression may not be directly expressed biologically as an immune-mediated disease but rather as a potential risk due to the reduced ability of the host to resist natural and acquired diseases. Animal models have been extremely valuable in identifying im- munotoxic agents and in developing immune profiles, identifying mechanisms of action, and alerting humans to potential health risks associated with exposure to specific xenobiotics, either consumed as drugs or through environmental exposure. Many immunoassays have been validated in animals to detect drug and chemical-induced im- munomodulation. Some of these bioassays are sensitive and predictive in assessing immune dysfunction. Many other immune assays that either are in the developmental stage or are not being used extensively are likely to have application in immunotoxicology. In general, these procedures will require additional testing and confirmation before they can be widely accepted as validated. Although many tests are available to screen for immunotoxicants, choice of an initial test that evaluates T-cell-dependent antibody response permits assessment of several compartments of the immune system concomitantly. These procedures are easily performed, quantitative, sensitive, and economi- cal, and they routinely detect a large percentage of the known im- munotoxic agents. A variety of tests are available to assess humoral and cellular im- munity as well as nonspecific resistance in humans. A series of tests can be performed in steps to assess immune competence in individuals who have been exposed to an immunotoxicant or potential immuno- toxicant. Some of these procedures parallel animal studies and require validation prospectively in populations exposed to putative immuno- toxicants and in control groups to ascertain their predictive value as

78 GUIDELINES FOR DEVELOPING SMACS immune-compromising agents. This aggressive approach will permit use of sensitive procedures for detection of immunomodulation in humans. RISK ASSESSMENT Noncarcinogenic Effects Toxicological risk assessment for agents without the capacity to induce carcinogenic or mutagenic effects has traditionally been based on the concept that an adverse health effect will not occur below a certain level of exposure, even if exposure continues over a lifetime. The existence of a so-called "threshold" dose is supported by the fact that the toxicity of many agents is manifest only after the depletion of a known physiological reserve. In addition, the biological repair capa- city of many organisms can accommodate a certain degree of damage by reversible toxic processes (Klaassen, 1986; Aldridge, 1986). Above the threshold dose, however, the homeostatic physiological processes that allow compensatory mechanisms to maintain normal biological function may be overwhelmed, leading to organ dysfunction. Thus, the objective of classical toxicological risk assessment is to establish a threshold dose below which adverse health effects are not expected to occur or are extremely unlikely. The concept of a no-effect level was introduced by Lehman and Fitzhugh (1954), who proposed that an acceptable daily intake (ADI) could be calculated for contaminants in human food. This concept was endorsed by the Joint FAO/WHO (Food and Agricultural Organization and World Health Organization) Expert Committee on Food Additives in 1961 and subsequently adopted by the Joint FAO/WHO Meeting of Experts on Pesticide Residues in 1962 (cf. McColl, 1990). Formally, an ADI is defined by the relationship ADI = NOEL/SF, where NOEL is the no-observed-effect level in toxicological studies (the highest experimental dose at which effects are not observed) and SF is the safety factor that allows for variations in sensitivity to the test agent in humans as compared with experimental animals and for

ESTABLISHMENT OF SMACS 79 variations within the human population. These two sources of varia- tion have often been accommodated through the use of a 10 x 10 = 100-fold safety factor reviewed by NRC's Food Protection Committee (NRC, 1970). In 1977, the NRC's Safe Drinking Water Committee reviewed the methods that had evolved for establishing ADIs and made several important recommendations. First, the committee proposed that NOEL be expressed in milligrams per kilogram of body weight rather than milligrams per kilogram of diet to adjust for dietary consumption patterns. Second, the committee suggested reducing the traditional 100-fold safety factor to only 10-fold in the presence of dose- response data derived from human studies. Third, the committee proposed augmenting the traditional 100-fold safety factor to 1,000- fold in the absence of adequate toxicity data (NRC, 1977). Although the use of safety factors is now accepted practice in es- tablishing exposures for noncarcinogenic effects, the NOEL/SF approach is subject to certain limitations (Munro and Krewski, 1981). Since ADI is only an estimate of the population's threshold dose, ab- solute assurance of safety is not provided (Crump, 1984a). Smaller experiments tend to yield larger NOELs, and hence larger ADIs, than larger, more sensitive experiments (Mantel and Schneiderman, 1975). Safety factors of 10-fold that are used to account for both inter- and intra-species variation in sensitivity cannot be guaranteed to provide adequate protection in all cases. For these reasons, ADI should not be viewed as possessing a high degree of mathematical precision but should be viewed as a guide to human exposure levels that are not expected to present serious health risks. In 1988, EPA recommended using the term "uncertainty factor" (UF) rather than safety factor in recognition of the uncertainty as- sociated with ADI and relabeled ADI as a reference dose (RfD) (Barnes and Dourson, 1988). EPA also introduced an additional modi- fying factor (MF) to account for specific scientific uncertainties in the experimental data used to establish RfD. The no-observed-adverse- effect level (NOAEL) is defined as the highest experimental dose at which no statistically significant increase in the occurrence of adverse effects is observed beyond that exhibited under control conditions. The RfD is determined using the relationship RfD = NOAEL/(UF x MF).

80 GUIDELINES FOR DEVELOPING SMACS In the present context, adverse effect is defined as any effect that contributes to the functional impairment of an organism or that reduces the ability of the organism to respond to additional challenges (Dourson, 1986). When the data do not demonstrate a NOAEL, a LOAEL (lowest-observed-adverse-effect level) may be used. A LOAEL is defined as the lowest experimental dose at which a statis- tically significant increase in the occurrence of adverse effects is ob- served. Five factors may contribute to the determination of the uncertainty and modifying factors. These are (1) the need to accommodate human response variability, including sensitive subgroups; (2) the need to extrapolate from animal exposure data to humans, when human ex- posure data are unavailable or inadequate; (3) the need to extrapolate from subchronic to chronic exposure data, when the latter are unavail- able; (4) the need to account for using a LOAEL, when a NOAEL is unavailable; and (5) the need to extrapolate from a data base that is inadequate or incomplete. Factors between 1- and 10-fold are often used to account for each of these sources of uncertainty. EPA has adapted the oral RfD methodology to estimate inhalation reference concentrations (RfCs) to be consistent in setting levels of noncarcinogenic chemicals (EPA, 1990). The inhalation RfC method- ology departs from the oral RfD paradigm by incorporating dosimetric adjustments to scale the exposure concentration for animals to a human equivalent concentration. For noncarcinogenic effects, the RfC approach should be the pri- mary method used for setting SMACs. Uncertainty (safety) factors between 1 and 10 should be used for each source of uncertainty listed above, depending on the nature and severity of the adverse effects. For setting SMACs, the duration extrapolation, if any, would be in the opposite direction from that indicated in the third source above, that is, it would be from long-term to short-term exposure. If an expo- sure-extrapolation approach such as Haber's rule is used and there is considerable uncertainty as to its validity, then the exposure extrapo- lation should be accompanied by the highest safety factor (10). In addition to the RfC method, alternative methods such as the bench- mark dose approach (Crump, 1984a; Chen and Kodell, 1989) should be considered as secondary approaches to setting SMACs.

ESTABLISHMENT OF SMACS 81 Carcinogenic Effects Mathematical Modeling For carcinogenic effects, particularly those considered to be due to genotoxic events such as alkylation of DNA, a threshold dose may not exist. Beginning with the pioneering work of Mantel and Bryan (1961), attempts have been made to estimate carcinogenic risks on a precise quantitative basis. This task has involved fitting mathematical models to experimental data and extrapolating from these models to predict risks at doses well below the experimental range. The probit-log dose model initially used by Mantel and Bryan (1961) was a carryover from the tolerance distribution models used for analyzing dose-response mortality data from acute toxicity studies. This model was later replaced by mathematical models based on presumed mechanisms of carcinogenesis. The mechanistic model used most frequently for low-dose extrapolation is the multistage model of Armitage and Doll (1961). A historical perspective on the evolution of the multistage model has been provided by Whittemore and Keller (1978). According to the multistage theory, a malignant cancer cell develops from a single stem cell as a result of a number of biological events (e.g., mutations) that must occur in a specific order. This model predicts that the age-specific cancer incidence rate should increase in proportion to age raised to a power related to the number of stages in the model. This model provides a good description of many forms of human cancer that allow for two to six stages in the carcinogenic process. Assuming that the rates of transition between stages in the multi- stage model are linearly related to dose, the dose-response curve for the multistage model is linear at low doses. Low-dose linearity is generally assumed for chemical carcinogens that act through direct interaction with genetic material. It is supported by theoretical con- siderations in carcinogenesis (Krewski et al., 1989a) and by observa- tions of the linearity of DNA binding with a number of chemical carcinogens at low doses (Lutz, 1990). When carcinogenesis occurs after toxic tissue injury, low-dose linearity may not be applicable, but the assumption is widely used in applications of low-dose risk assess- ment for regulations in the absence of clear information to the con- trary (OSTP, 1985).

82 GUIDELINES FOR DEVELOPING SMACS EPA (1986) uses the linearized multistage model for low-dose ex- trapolation. With this procedure, an upper confidence limit, QJ*, on the coefficient of the linear term in the model is obtained (Crump, 1984b). The value qf represents an upper bound on the slope of the dose-response curve in the low-dose region and on the excess risk above background associated with a unit measure of dose (such as 1 mg/kg of body weight per day). Only an upper confidence limit on the low-dose slope is used by EPA for extrapolation, in part because point estimates of this parameter are highly unstable. Linear extrapo- lation from an upper confidence limit on excess risk obtained in the experimental range is generally regarded as conservative in the sense of protecting human health. It will be conservative as long as the dose-response relationship in the low-dose region is convex, regardless of its actual mathematical formulation (Gaylor and Kodell, 1980). Recently, a two-stage model that incorporates tissue growth and cellular kinetics has been proposed as an alternative to the multistage model for cancer risk assessment (Moolgavkar and Venzon, 1979; Moolgavkar and Knudson, 1981). Mathematical formulations of this stochastic birth-death-mutation model have been given by Moolgavkar and Luebeck (1990) and by Greenfield et al. (1984). The model assumes that two genetic events (mutations), each occurring at the time of cell division, are necessary for a normal cell to become malignant. Allowance is made for natural growth of the target tissue and for clonal expansion of the pool of cells that have undergone the first mutation. The birth-death-mutation model provides a convenient framework for describing initiation-promotion-progression mechanisms of car- cinogenesis. Initiating activity may be quantified in terms of the rate of occurrence of the first mutation. The rate of occurrence of the second mutation describes progression to a fully differentiated can- cerous lesion. Promotional activity is measured by the difference in the birth and death rates of initiated cells. Carcinogenic agents may be classified as initiators, promoters, or progressors according to com- ponents of the model that they are presumed to affect (Thorslund et al., 1987). Application of the two-stage model with clonal expansion in risk assessment for putative tumor-promoting agents has been investigated on a limited basis (Thorslund and Charnley, 1988). Before this model can be recommended for routine application, its statistical properties require further study, particularly with respect to predictions of risk at low doses (Portier, 1987). In the absence of promotional effects and

ESTABLISHMENT OF SMACS 83 variability in the pool of normal cells, the two-stage birth-death- mutation model is reduced to the two-stage version of the classical multistage model. Pharmacokinetic Considerations Chemical carcinogens may require some form of metabolic activa- tion to exert their effects. If metabolic activation can be characterized adequately in terms of a pharmacokinetic model, then the dose deliv- ered to the target should be used in place of the administered dose for purposes of dose-response modeling and low-dose extrapolation (Hoel et al., 1983). In general, the use of delivered rather than administered doses may be expected to lead to more accurate predictions of car- cinogenic risk (NRC, 1987b). Such predictions, however, may intro- duce substantial uncertainty if complicated, physiologically based pharmacokinetic models with 20 to 40 parameters, each subject to uncertainty, are used for tissue dosimetry (Portier and Kaplan, 1989; Farrar et al., 1989). Intermittent Exposure Most bioassay data suitable for low-dose extrapolation reflect con- tinuous exposure to a carcinogen, necessitating the translation of risks calculated for continuous exposures to risks associated with short-term or intermittent exposures. Crump and Howe (1984) developed a meth- odology for applying the multistage model to carcinogenic risk assess- ment when exposure to a carcinogen is intermittent. Their results have been adapted by the Committee on Toxicology (COT) (NRC, 1986b) for setting emergency and short-term exposure guidelines for chemicals of interest to the U.S. Department of Defense. Both multi- stage and two-stage models have been studied with respect to age at which exposure occurs (Kodell et al., 1987; Chen et al., 1988; Murdoch and Krewski, 1988; Krewski and Murdoch, 1990). The results of these investigations indicate that when an early stage in the car- cinogenic process is dose-dependent, early exposures will be of greater concern than later exposures. Conversely, when a late stage is dose- dependent, late-life exposures pose a higher risk. In the context of the classical multistage model, the potential increased excess risk from

84 GUIDELINES FOR DEVELOPING SMACS exposure for a fraction, /, of a lifetime at a given dose rate, d, will never be more than k times the excess risk from full lifetime exposure at dose rate fd, where k is the number of stages in the model. (Note that administration of daily dose d for fraction / of a lifetime yields the same cumulative lifetime dose as administration of dose fd over the course of a lifetime.) With the two-stage model, however, the increased risk can be substantial when the agent of interest greatly increases the proliferation rate of the initiated cell population. Carcinogenic Mixtures The interactive effects between two carcinogens have been exten- sively studied (Krewski and Thomas, 1992). However, the existence of synergism at low levels of exposure cannot be assessed directly. The multistage and the multiplicative relative-risk models for cancer predict near additivity of excess risk at low doses, even when there is nonadditivity at higher doses (NRC, 1988c; Krewski et al., 1989b; Chen et al., 1990). Kodell et al. (1991) recently described departures from additivity within the context of the two-stage birth-death- initiation model. Additivity is expected, however, at low doses for initiators and progressors, as well as nongenotoxic promoters that may demonstrate a threshold. In the absence of data from experiments involving joint exposure to multiple agents, the health risk from joint exposure to two or more carcinogens must be based on the results of experiments with single compounds. In the absence of evidence to the contrary, the total risk from a mixture of carcinogens generally is assumed to equal the sum of the individual risks of the components (EPA, 1986). This assump- tion is consistent with the concept of a group-limit value, which implies that individual allowable concentrations must be reduced when they are in a mixture of chemical toxicants with the same or similar effects to control the overall risk. Chen et al. (1990) noted that if upper confidence limits on individual estimates of carcinogenic risk of the same order of magnitude simply are summed as an approxima- tion to an upper confidence limit on the total excess risk of a mixture, the upper bound will be conservative. Using the assumption of additivity of component risks to overcome this conservatism, Chen et al. (1990) developed a formal procedure for calculating an upper con- fidence limit on the total carcinogenic risk of a mixture.

ESTABLISHMENT OF SMACS 85 Interspecies Extrapolation Quantitative methods for cancer risk assessment for carcinogenicity must also take into account species differences. Traditionally, dose equivalency among species has been based on body weight. However, many physiological constants (e.g., consumption of water, food, and oxygen) vary as a power function of body weight (NRC, 1987c). Dourson and Stara (1983) pointed out that toxicity of many com- pounds may be related more closely to body surface area than to body weight (Pinkel, 1958; Freireich et al., 1966). The use of a general power function of body weight for species conversion includes body- weight conversion as a special case with a power of unity. Because body surface area varies approximately in proportion to the 2/3 power of body weight, surface-area conversion is another special case of the general power-law relation. Based on a reanalysis of the data of Freireich et al. (1966), Travis and White (1988) suggested that body weight to the 3/4 rather than to the 2/3 power may be more appro- priate for species conversion. Allen et al. (1988) found a high cor- relation of carcinogenic potency in animals and humans but were unable to resolve whether body weight or surface area provided a better basis for species conversion. If the compound is well defined in its metabolism and phar- macokinetics, these data should be used in preference to the body- weight to surface-area correction. The body-weight to surface-area correction is essentially a correction for metabolic rate based on body size. The basis of this correction is more related to rates of basal metabolism than of xenobiotic metabolism. Thus, in the face of more specific data on the comparative pharmacokinetics and metabolism of the chemical in question, the correction should not be used. Extrapolation of Animal Data to Humans Interpretation of data derived from animal experiments requires experienced scientific judgment in a variety of disciplines. Consider- ation should be given to the conditions under which the data were obtained and their relevance to conditions of human exposure under question. How similar are the test species and the test organ to humans and the corresponding human organ in morphology, in sen- sitivity of response to the contaminant, and in metabolism and dispos- ition of the contaminant? Were the observed responses of animals the

86 GUIDELINES FOR DEVELOPING SMACS consequences of exposure conditions to which humans may be sub- jected? Development of SMACs requires that animal data be extrapolated to humans. The species most representative of humans, considering both toxicological and pharmacokinetic characteristics, should be used for determining the appropriate exposure limit. If data are not avail- able on which species best represents humans, it is prudent to use data from the most sensitive animal model to set appropriate limits. As discussed earlier, uncertainty factors can be used when an es- timate of the NOAEL in animals is available. In the application of uncertainty factors, particularly susceptible individuals should be considered. People at unusual risk might include those at risk because of unusual physical exertion and other stresses; such risk might occur to astronauts. Persons of suspected high susceptibility (i.e., very old or those with known diseases) are unlikely to be space-station work- ers, thus reducing the need for safety factors to allow for these states or conditions. The value of the uncertainty factor also will depend on the quality of the animal data used to determine the NOAEL. The RfD often has been set at a level 100-fold below the NOAEL derived from lifetime animal experiments. Larger uncertainty factors also have been used; for example, a 1,000-fold factor usually is used to establish an RfD when only subchronic (90-day) toxicity studies are available. Higher uncertainty factors also may be used because of the altered physiological status of astronauts in the space shuttle or space station. Conversion from animals to humans may be done on a body-weight or surface-area basis, as discussed earlier. When available, pharmaco- kinetic data on tissue doses also may be considered for use in species conversion. In using toxicological data, data may have to be extrapo- lated from those on oral exposure of animals (in milligrams per kilo- gram of body weight) and converted to inhalation exposure of humans (in milligrams per cubic meter of air). Two approaches are possible: extrapolation from rat oral data to human oral dose and then to human inhalation concentration; or extrapolation from rat oral data to rat inhalation concentration and then to human inhalation concentration. The first approach does not assume similar breathing rates for animals and humans. A 70-kg man breathes air at 7-10 L/min at rest, 20 L/min when moderately active, and 40-60 L/min when engaged in heavy work. A resting 250-g (0.250-kg) rat breathes air at 200 ml/min (0.2 L/min).

ESTABLISHMENT OF SMACS 87 To illustrate the first approach, assume that an oral dose of 1 mg/kg of body weight per day in rats is equivalent to an oral dose of 1 mg/kg of body weight per day in humans. A 70-kg man at rest inhales 15m3 of air over a 24-hr period. The concentration inhaled by a 70-kg man in 1 day is shown by the equation [(1 mg/kg)(70 kg)]/(15 m3) = 4.7 mg/ms. However, not all of what is inhaled will be absorbed. Toxi- cokinetic information available on the uptake of the compound of interest or on similar compounds should be used to correct the above equation. In the above example, if only 50% of an inhaled concentra- tion of a material is deposited or absorbed by humans, then an indi- vidual would have to be exposed to a concentration of 9.4 mg/m3 over a 24-hr period to achieve the equivalent of an oral dose of 1 mg/kg per day. To illustrate the second approach, suppose that a 250-g rat at rest inhales air at 200 ml/min or 0.288 m3 over a 24-hr period. An oral dose of 1 mg/kg in the rat is equivalent to an inhalation concentration in the rat, as shown by [(1 mg/kg)(0.250 kg)]/(0.288 m3) = 0.87 mg/m3. For a 70-kg man inhaling 15 m3 of air in 24 hr, the conver- sion to an inhalation concentration is (0.87 mg/m3)(0.288 ms/0.250 kg)(70 kg/15 m3) = 4.7 mg/m3. Again, corrections should be made for the percentage of inhaled material deposited or absorbed. SMACs for Carcinogens The determination of SMACs for carcinogens generally will require extrapolation of toxicological or epidemiological data obtained under conditions of long-term (often lifetime) exposure to periods of shorter duration, such as 1, 30, or 180 days. Here, we illustrate how the methods proposed by Kodell et al. (1987) based on the multistage model can be used to set SMACs for carcinogens. Suppose that data from a long-term animal bioassay or human epi- demiological study have been used to arrive at an average daily life- time exposure level, d, of a particular chemical, corresponding to a particular lifetime excess carcinogenic risk. As reported by COT, a lifetime risk level of 10"4 has been used by the U.S. Department of Defense (NRC, 1986b) to set exposure guidelines for carcinogenic chemicals. Assume that d has been estimated with the use of any available pharmacokinetic information to determine tissue dose, the

88 GUIDELINES FOR DEVELOPING SMACS linearized multistage model, and an appropriate interspecies conver- sion factor. Suppose that the carcinogen of interest affects only a single stage of a multistage process and that there are up to k = 6 stages in total. Let i (i =1,2, ..., k) denote the index of the stage that is dose-related. Consider the case of an astronaut exposed to a constant daily dose, D, from age /0 to t1. Using results of Kodell et al. (1987, p. 340), it can be shown that exposure to a dose, ' from time tQ to time /x (0 < t0 < /j < /) will yield the same excess risk as a constant daily dose, d, from birth to age /, where / denotes the time when excess risk is evaluated (usually the average human life- time). To illustrate, consider a three-stage process with only the first stage dose-related, and suppose that exposure occurs at the earliest age possible. The assumed earliest age of exposure for astronauts will be age 25 (9,125 days), and the assumed average human lifetime will be 70 years (25,550 days). To determine the 180-day SMAC, given the acceptable lifetime daily dose d, the value of D in Eq. 1 is evaluated with k = 3, t = 25,550, t0 = 9,125, and t1 = 9,125 + 180 = 9,305. This gives D = 115.8 d. For a 30-day SMAC, k = 3, t = 25,550, /„ = 9,125, and /j = 9,155. Substituting these values into Eq. 1 gives D = 688.2 d. An alternative but equivalent approach is to calculate an intermedi- ate dose, £>*, that would yield the same total exposure as the average daily lifetime exposure level d, and then apply an adjustment factor, /, based on the multistage model. This approach has been followed by COT (NRC, 1986b) in performing risk assessments for the U.S. De- partment of Defense. With this approach, - <o>. (2) which is simply the average daily dose from age t0 to /j that wi_ ^fc yield the same cumulative dose as continuous exposure for a li

ESTABLISHMENT OF SMACS 89 to a daily dose of d. The SMAC is then defined by D*/f, where (3) In the preceding example, D* would be 141.9 d for the 180-day SMAC and 851.7 d for the 30-day SMAC. The value of / would be 1.23 for the 180-day SMAC and 1.24 for the 30-day SMAC. Con- sidering both the multistage and two-stage models, Murdoch et al. (in press) suggest that the value of / for astronauts between 25 and 45 years of age is unlikely to exceed a value of »2. The case of near instantaneous exposure (e.g., single-day exposure) at time /0 to a total dose D requires separate treatment. In this limiting case, SMACs for exposure durations of 1 day or less may be based on the relationship dtk (4) which applies in the case of a carcinogen that affects only the /'* stage of a Ac-stage process. The corresponding adjustment factor for the time-weighted average dose, D* = dt, is {£(,!,) [(*-;+ f — i * J '- (/•-I) (5) It must be remembered that extrapolation from a daily lifetime ex- posure level and conversion to an instantaneous exposure level using Eq. 4 or Eq. 5 is an extreme case and is valid only under the assump- tions underlying the multistage theory of carcinogenesis. As in Eq. 3, the value of / based on Eq. 5 is unlikely to exceed 2 (Murdoch et al., in press).

90 GUIDELINES FOR DEVELOPING SMACS Use of OSHA, ACGIH, NRC, or Other Limits There are several inhalation guidelines that can be used as sources of information on compounds of interest. The guidelines, established by OSHA,a ACGIH,b EPA, ATSDR,C and NRC, should be reviewed before establishing SMACs. The subcommittee has reviewed reports of analyses of atmospheric contaminants detected in a variety of closed environments: seven manned spaceflights (Mercury and Gemini series), many space-shuttle flights, several ground-based simulated cabin atmospheres (both manned and unmanned), nuclear submarine and Sealab experience, and analyses of off-gas products from cabin materials. Nearly 300 compounds that could be possible space-station contaminants have been identified. Many of these compounds have been reviewed by the NRC's COT, and 1 - and 24-hr emergency exposure guidance levels (EEGLs) or 90- day continuous exposure guidance levels (CEGLs) for submarines have been recommended. The recommended EEGLs are similar to 1 - and 24-hr SMACs and permit some reversible effects, such as headache and mild irritation. The differences between the conditions for purg- ing the atmospheres of spacecraft and of submarines must be con- sidered. Although designing for a complete atmospheric purge by opening the spacecraft may be possible, the need for such an emer- gency operation must be minimized. In contrast, submarines can sur- face to ventilate except in extraordinary circumstances. EEGLs differ from STELs (short-term exposure limits) recom- mended by OSHA or ACGIH in that STELs are generally 15-min limits to which workers may be exposed daily for many years. The 1 - and 24-hr SMACs, on the other hand, are limited for infrequent emergency exposure. The continuous exposure guidance levels (CEGLs) recommended by COT for submarines are ceiling concentrations designed to avoid ad- verse health effects, either immediate or delayed, of more prolonged exposures and to avoid degradation in crew performance that might endanger the objectives of a particular mission as a consequence of continuous exposure for up to 90 days. *Occupational Safety and Health Administration. bAmerican Conference of Governmental Industrial Hygienists. °Agency for Toxic Substances and Disease Registry.

ESTABLISHMENT OF SMACS 91 ATSDR is developing documents on several industrial chemicals. These documents provide detailed background information on the chemicals and the basis for setting limits for those chemicals. EPA has developed numerous cancer assessments and inhalation RfCs that are accessible on EPA's Integrated Risk Information System (IRIS). Both values assume a 70-year exposure, which would require adjustment to the space scenario. EPA's cancer assessment methodol- ogy has been described in The Risk Assessment Guidelines of 1986 (EPA, 1987). All the documents used to establish previous industrial or public exposure limits for airborne materials should be reviewed for perti- nent information before establishing SMACs. The purpose of this comparison is not to mimic the regulatory levels set by others, but to determine if the SMACs are reasonable in light of the special needs of NASA. The background information on how other agencies set guid- ance levels also should be useful in setting SMACs. Exposure to Mixtures Individual spacecraft limits for single compounds have been estab- lished for guidance in engineering design of atmosphere-handling systems and represent an upper limit for each compound without re- gard to its potential occurrence in mixtures or in the presence of other toxicants. However, atmospheric contaminants are encountered most often as complex mixtures, and the toxicological hazard to humans from inhalation, especially on a long-term basis, must be assessed in terms of permissible atmospheric loadings for these mixtures. There- fore, individual limits must be integrated into sets of group limits to reflect air-quality conditions judged to be safe for humans during given exposure periods. The spacecraft atmosphere consists of a mixture of compounds, many of which have similar effects that are likely to be additive. These potential additive effects must be considered in the assessment of the toxicological hazard of contaminant mixtures. The following guidelines for SMAC development provide for the potential summa- tion of toxic effects of contaminants. 1. The concentration of each contaminant in the spacecraft atmos- phere must not exceed its SMAC (1 hr, 24 hr, or 180 days) value

92 GUIDELINES FOR DEVELOPING SMACS expressed as a time-weighted average or as a ceiling value, as appro- priate. 2. For each one of the groups of contaminants, a group-limit con- cept will be utilized to evaluate the toxicological hazard of the group. In each group, the summation, T, of the ratios of concentration to the SMAC value of each member of the group must not exceed unity (ACGIH, 1989). The following formula will apply when the compo- nents in a mixture from structurally related compounds have similar toxicological effects: Ci Cz C3 • C" - r* 1, (6) SMAC2 SMAC} SMACn where Ct is the observed concentration of the /* material and SMAC, is the corresponding permissible concentration. Example: Air contains 400 ppm of acetone (SMAC, 750 ppm), 150 ppm of sec-butyl acetate (SMAC, 200 ppm), and 100 ppm of methyl ethyl ketone (SMAC, 200 ppm). 400 + i50 + WO _ + + _ 750 200 200 Although each individual exposure does not exceed its SMAC, the permissible total is exceeded, and exposures must be reduced (ACGIH, 1989). There will be situations in which the effects of exposure to mixtures will be greater than additive, and this information must be taken into account in setting SMACs. Physiological adaptive responses that are important in space opera- tion may involve muscle remodeling, loss of red-blood-cell mass (anemia), altered nutritional requirements, and behavioral changes. The panoply of all physiological changes considered important demon- strates that the individual astronaut is in an altered homeostatic state. How this altered state modifies reactions to chemicals in the space - station environment requires additional information. With current knowledge, the maximum allowable concentrations of airborne chemi- cals in the closed system of space travel would have to be reduced from permissible levels on earth. The astronaut in space is an atypical human, particularly following extended residence in space.

ESTABLISHMENT OF SMACS 93 Some spacecraft contaminants may present a carcinogenic risk. In this case, SMACs should satisfy the following guidelines: 1. The lifetime excess risk of cancer resulting from exposure to a chemical contaminant in space for periods up to 180 days should not exceed 10~4 or other levels acceptable to NASA. The lifetime risk associated with such short-term exposures may be estimated on the basis of a time-weighted average lifetime dose in Eq. 2, divided by an adjustment factor of / = 2. 2. The lifetime risk for joint exposure to two or more carcinogens should not exceed 10~4 or other levels acceptable to NASA. Assuming additivity of risks, the total risk is determined by summing the indi- vidual risks. Modification of Contaminant Toxicity by Environmental Factors The special conditions of the space environment must be taken into account in defining spacecraft contaminant standards. Examples of potential interaction between contaminants and environmental factors include the effects of increased levels of radiation on the radiomimetic action of benzene and the effects of increased fatigue from stress factors (e.g., noise, vibration, and weightlessness (Kaplan, 1979)). Lung function may be different, deposition of particles is different under microgravity conditions, the potential toxic effects of inhaled particles on CNS active agents may be different, and the toxicity of gases may be altered by condensation on particulates in the spacecraft. Astronauts will be physically, physiologically, and psychologically compromised for the following reasons: loss of muscle mass, loss of bone mass, depressed immune system, decreased red-blood-cell mass (anemia), altered nutritional requirements, behavioral changes from stress, fluid shift in the body, altered hormonal status, and altered drug metabolism. These changes imply that an astronaut in space will be in an altered homeostatic state and may be more susceptible to toxic chemicals. The adaptive physiological changes that occur in microgravity may prove to be detrimental to astronauts. For example, Merrill et al. (1990) analyzed a number of liver enzymes from rats that had flown aboard Cosmos 1887. They found a decrease in amounts of cyto- chrome P-450. In addition, decreases occurred in the enzyme aniline hydroxylase and ethylmorphine Af-demethylase. Since this family of

94 GUIDELINES FOR DEVELOPING SMACS enzymes is responsible for the metabolism of steroid hormones and a variety of xenobiotics, including antibiotics and drugs, this observa- tion could have important implications for space medicine and toxi- cology. The changes in drug-metabolizing enzymes, immunological changes, and other alterations could be important factors in disease susceptibil- ity and chemical toxicity, and those should be taken into account to recommend SMACs. Risk assessment of chemicals is based on biological studies conduct- ed in animals and humans on earth. It is unclear how much of the data can be generalized to the space environment. A variety of as- sumptions will be necessary. The quality of these assumptions can be improved substantially as more precise human data are gained from astronauts. Therefore, it is important to take into account the physio- logical changes induced in the space crew and the impact of these changes on SMAC values for various contaminants. GENERAL APPROACH TO ESTABLISHING SMACs The first step in producing a document describing the SMAC for a chemical is to collect and review all relevant information available on the compound. Based on the review of the literature, SMACs for different time periods are developed and a rationale is provided for each recommendation. The next step is to review and evaluate prior EEGLs and CEGLs, ACGIH's threshold limit values (TLVs), OSHA's recommendation for permissible exposure limits (PELs), NIOSH's (National Institute of Safety and Health) recommended exposure limits (RELs), ATSDR's recommendations, and EPA's ambient air quality standards. The intent of such a review is not to mimic the regulatory levels set by others but to make use of all available background infor- mation gathered by other agencies in evaluating the health effects of the chemical of interest. Development of 1 - and 24-hr SMACs for different durations of ex- posure usually begins with providing a SMAC for the shortest expo- sure of 1 hr. Values for 24-hr SMACs typically are developed by using Haber's law when applicable. Important determinants of the nature and severity of biological effects resulting from exposure to a chemical are the level and duration of exposure. There has been a general interest in assessing the extent to which the concentration, C, of a given chemical and the duration of exposure, T, interrelate to

ESTABLISHMENTOFSMACS 95 determine the magnitude of the biological response, K. The overall relationship between C, T, and K is a complex, three-dimensional surface. However, if interest is in a particular level of biological response, for example, K = k, the problem becomes two-dimensional. By restricting the biological response to a particular level, k, Haber postulated that the formula CxT = k could be used to relate the toxic effect of certain inhalable substances to their concentration and time of exposure. Although the above formula, which sometimes is re- ferred to as Haber's law, is clearly valid only for a limited number of substances and only for certain combinations of concentration and exposure time, it has been used to synthesize data relating concentra- tion and time of exposure. A more general expression for examining relationships between C, T, and k is given by C" x Tb = k, where the exponents a and b are estimated from the data. This formula allows for the fact that C and T do not always contribute equally to the ob- served toxicity. When detoxification or recovery occurs and data are available on 24-hr exposures, these factors must be taken into account in modify- ing Haber's law (C x T = k). Haber's law is inappropriate for materi- als, such as ammonia and NO2, that have been shown to be more toxic with high concentrations over short periods. High concentrations of particulates also have been shown to overload normal host defense systems, making predictions from such data bases unreliable for long- term, low-level exposures. Consequently, caution must be exercised in using even the general expression for C, T, and k when evaluating exposure conditions that are likely to result in comparable effects in laboratory animals and humans. The approach to establishing 1- or 24-hr SMACs is judged on a substance-by-substance basis. For substances that affect several organ systems or have multiple effects, all end points—including reproductive (in both sexes), developmental, carcinogenic, neurotoxic, respiratory, and other organ-related effects—are evaluated, the most important or most sensitive effects receiving the major attention. For the many compounds, such as upper respiratory tract irritants, for which there are reliable human data and a high degree of confidence that effects from a single exposure will be reversible, then NASA may establish exposure limits directly from the available data. For those compounds that have few human data but sound animal data and effects that are likely to be reversible following a single exposure, a species extrapola- tion factor may be employed, the magnitude of which will depend on the quality of the data.

96 GUIDELINES FOR DEVELOPING SMACS In the absence of better information, a safety factor of 10 is sug- gested for 1- and 24-hr SMACs when only animal data are available and extrapolation from animals to humans is necessary for acute, short-term effects or when the likely route of human exposure differs from that of a relevant experiment. With carcinogenic chemicals, the SMAC is set so that the estimated risk of a neoplasm in the upper 95% confidence limit is no more than 1 in 104 or some other risk acceptable to NASA. SMACs do not represent a sharp dividing line between safe and unsafe concentrations. Some people are expected to be affected ad- versely when SMACs are exceeded. However, some persons may be affected adversely even when SMACs are not exceeded. When expo- sure lasts longer than 24 hr, the 24-hr SMAC no longer applies, and appropriate measures must be taken to comply with the concentration implied by the corresponding 7-day, 30-day, and 180-day SMAC values. The 30- and 180-day SMACs are generally 0.01-0.1 times the 24-hr SMAC for noncarcinogenic substances. When detoxification is sub- stantial, a division by 10 may be more appropriate. When detoxifica- tion is not evident or is slow, a division by 100 or more may be more appropriate. When the substance accumulates in tissues, as do halo- genated biphenyls and metals, even larger safety factors may be used. In other cases, lower safety factors may be applied, depending on the data. The choice within these general guidelines must be determined for each material separately. When data from chronic studies are available, they are used to derive 180-day SMACs, applying safety factors as needed. With carcinogenic chemicals, an estimate of risk is provided for the recommended SMACs. When a substance is known to cause an effect that will be aggravated by microgravity, additional safety factors should be used. The suggested format for a SMAC document is shown in Appendix 1. The conversion factors and the reference values used in establishing SMACs are listed in Appendixes 2 and 3, respectively.