Click for next page ( 18


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 17
Markers of Exposure In all matters of toxicologic concern, one must relate the effect of a toxicant to the dose required to cause the effect. Our ability to determine dose accurately varies widely with the kind of study in- volved-from epidemiologic studies, in which dose estimation must often be based on limited information from area monitor- ing, to studies performed in vitro, in which the dose to a subcellular organelle can be measured exactly. One way to obtain an accurate measure of dose is to measure biologic markers of exposure. A biologic marker of exposure is a substance measured in a compartment within an organism-an exogenous substance or its metabolite or a product of an interaction between a xeno- biotic agent and some target molecule or cell. This chapter examines biologic markers of exposure. Exposure is the sum of xenobiotic materi- al presented to an organism, whereas dose is the amount of the material that is ab- sorbed into the organism (internal dose) or the amount of active material that reaches the site of toxic action (in the case of biologically effective dose). Thus, there is an important distinction between exposure and dose that should be considered in the assessment of human ex- posure to toxic substances. In many animal toxicologic studies and human clinical studies, the internal dose i' s inferred from knowledge of the exposure to a toxicant administered by ingestion or injection. However, if the toxicant is volatile, much of the compound might be exhaled unchanged soon after ingestion or intraperitoneal injection, and the re- tained dose could be less than the adminis- tered dose. If the toxicant is delivered by inhalation, the internal dose is dif- ficult to assess and will depend on expo- sure conditions (air concentration, tim- ing of exposure, etc.), deposition and absorption efficiencies of the inhaled material, metabolism, and ventilation patterns of the exposed subject. In inhalation exposures, the estimated dose is sometimes expressed as duration of exposure to a toxicant at a given atmos- pheric concentration. It is particularly important for regulatory agencies to be able to relate the appearance of an adverse health effect to specific atmospheric concentrations of a toxicant, because regulations are usually set in terms of allowable concentrations in air. In the absence of toxicokinetic data, however, atmospheric concentrations used in toxi- cologic studies can be misleading, if the internal dose is not linearly related to the atmospheric concentration, especially at the high concentrations often used in such studies. If the mechanism of a toxic effect is 17

OCR for page 17
lg known (the ideal situation), one can speak of the biologically effective dose at the cellular or subcellular site of action; that is the actual dose causing the effect, but it might be difficult to measure, ex- cept in studies performed in vitro. Recent advances in molecular epidemiology (Per- era and Weinstein, 1982; Wogan and Gore- lick, 1985; Wogan, 1988) have provided techniques that allow sampling of such biologic markers as covalent adducts formed by the binding of toxicants with macromolecules at or near the site of toxic action. Such adducts could prove to be valuable markers of biologically effec- tive dose (e.g., DNA and protamine adducts) or of exposure (e.g., hemoglobin adducts). All those assessments of dose-i.e., biologically effective dose, dose to crit- ical tissues, internal dose, and atmos- pheric concentration-are required for various aspects of health-effects stud- ies, and biologic markers of each would be useful. Most important, however, is information on how each type of dose meas- urement is related to the others, so that measurement of some can be used to estimate others. Mathematical models based on the physical and chemical properties of the toxicant and the recipient organism have proved useful for extrapolation not only between different dose measurements, but also between species and between exposure regimens (NRC, 1987~. The mechanism of action of a toxicant is rarely known, and the dose to tissues where the toxic effects occur is usually the most useful expression of dose. Toxi- cokinetic studies are required to estab- lish the relationship between an adminis- tered dose and the dose to the tissue of concern. It is useful if the relationship between the dose to a target tissue and the appearance of an indicator substance in readily sampled body fluids can be estab- lished, so that the dose to a target tissue can be estimated from available biologic material. In the following sections, we first dis- cuss the factors that govern the deposition of inhaled materials-particles, gases, and vapors, those are the factors that determine the initial internal dose re- sulting from inhalation exposure. We then AI4RKERS IN PULMONARY TOXICOLOGY discuss the toxicokinetics of the depos- ited material, i.e., the rate and extent of clearance of deposited material and its metabolites from the respiratory tract, of their distribution in the body, and of their excretion. We then treat meth- ods and sites for monitoring for inhaled materials and their products in the body. All that information determines the dose to the critical tissue, or, if the mechan- ism of injury is known, can even reveal the biologically effective dose. Mathematical models used to extrapolate animal toxico- kinetic data to humans are discussed, and the use of clinical techniques to assess markers of exposure and dose is reviewed at the end of this chapter. DEPOSITION OF INHALED MATERIAL IN THE RESPIRATORY TRACT Particles Particulate matter can include solid, relatively insoluble particles and liquid droplets that can be readily soluble in body fluids. The deposition of both types of particles are governed by the same for- ces, but the disposition of the deposited material will depend on the chemical prop erties of the material. This section deals with the deposition of both types of par- ticles. ticles and the clearance of lipid-soluble material, whether inhaled as an aerosol or as a gas or vapor, is dealt with later. Inhaled particles can come into contact with airway surfaces and be deposited on them. The extent and site of deposition depend on various factors, such as par- ticle characteristics, ventilation pat- tern, and airway structure. Deposition can occur by five basic physical mechan- isms: impaction, sedimentation, Brownian diffusion, interception, and electrostat- ic precipitation. The clearance of insoluble par Im paction is inertial deposition. It occurs when a particle's momentum pre- vents it from changing course in an area where there is a change in the direction of airflow. Impaction is the main mechan- ism by which a particle having an aerody

OCR for page 17
EXPOSURE namic equivalent diameter (Dee) of 0.5 ,um or more is deposited in the upper res- piratory tract and at or near tracheobron- chial-tree branching points. The proba- bility of impaction is proportional to air velocity, rate of breathing, and par- ticle density and size. Sed imentation is deposition due to gravity. When the gravitational force on an airborne particle is balanced by the total of forces due to air buoyancy and air resistance, the particle will fall out of the airstream at a constant rate, known as the terminal settling velocity. The probability of sedimentation is propor- tional to particle residence time in the airway and to particle size and density and is inversely proportional to breathing rate. Sedimentation is an important depo- sition mechanism for particles with Dae of 0.5 ,um or more, which penetrate to air- ways whose air velocity is relatively low, e.g., mid-size to small bronchi and bronchioles. Submicrometer-size particles, espe- cially those with physical diameters of 0.2 Em or less, acquire a random motion due to bombardment by surrounding air mole- cules; this motion can result in contact with an airway wall. The displacement sustained by a particle is a function of the diffusion coefficient, which is in- versely related to particle cross-sec- tional area. Brownian diffusion is a major deposition mechanism in airways whose airflow is low or absent, e.g., bron- chioles and alveoli. However, extremely small particles can be deposited by dif- fusion in the upper respiratory tract, trachea, and larger bronchi. Interception is an important mechan- ism of deposition of fibers and occurs when a fiber edge makes contact with an airway wall. The probability of intercep- tion increases as airway diameter de- creases; fibers that are long and thin can penetrate into distal airways before deposition. Fiber shape is also impor- tant, in that straight fibers penetrate more distally than do curved fibers. In electrostatic precipitation, some fresh- ly generated particles are electrically charged and exhibit greater deposition than that expected from size alone. That 19 is due either to ionic charges induced on the surface of an airway by the par- ticles or to space-charge effects; repul- sion of particles bearing like charges results in increased migration toward the airway wall. The effect of charge on deposition is inversely proportional to particle size and airflow rate. Most am- bient particles become neutralized natur- ally because of the presence of air ions, so electrostatic deposition is generally a minor contributor to particle collec- tion by the respiratory tract; however, it is important in some laboratory studies. Patterns of deposition efficiency (i.e., percentage deposition of amount inhaled) in the human respiratory tract are shown in Figures 2-1 through 2-4. The use of different experimental methods and protocols results in considerable variation in reported values. Figure 2-1 shows the pattern for overall respira- tory tract deposition. Note the deposi- tion minimum over the size range 0.2-0.5 ,um. As previously discussed, particles with diameters of 0.5 Am or more are sub- ject to impaction and sedimentation, whereas the deposition of those 0.2 Am or less is diffusion-dominated. Par- ticles with diameters between these val- ues are minimally influenced by all three mechanisms and tend to have relatively long suspension times in air. They undergo minimal deposition after inhalation, and most are carried out of the respiratory tract in exhaled air. The effect of breathing mode on particle deposition in humans is evident from Fig- ure 2-1. Nasal inhalation results in greater total deposition of particles with diameters over 0.5 Am than does oral inhalation, because collection in the upper respiratory tract is greater. But there is little apparent difference in total deposition of particles of 0.2-0.5 ,um between nasal and oral breathing. Figure 2-2 shows the pattern of deposi- tion in the upper respiratory tract (the larynx and airways above it). Again, it is evident that nasal inhalation results in much greater deposition than oral. The greater the deposition of a substance in

OCR for page 17
20 loo 90 80 70 60 z lo- So - o 40 30 20 10 O _ Oral Inhalation Nasal Inhalation 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 11 1 1 1 1 1 1 ~ T . 1 ~ ~ I T T TT;I 1- . 0.01 0.1 1.0 PARTICLE DIAMETER (pm) FIGURE 2-1 Particle deposition efficiency in human respiratory tract. Values are means with standard deviations. Source: Reprinted with permission from Schlesinger, 1989. 100 90 80 70 ~ 60 - o 50 - O 40 ILL 30 20 10 PARTICLE DIAMETER (um) FIGURE 2-2 Particle deposition efficiency in human upper respiratory tract. Values are means with standard deviations. Source: Reprinted with permission from Schlesinger, 1989. MARKERS IN PULMONARY TOXICOLOGY 00 90 80 70 60 ~0 40 30 20 10 10 ~ Oral Inhalation Nasal Inhalation . 1~1 o . T] 11~7 ~ 0.01 0.1 1.0 10 O 00 90 80 70 60 50 40 30 20 10 o

OCR for page 17
EXPOSURE the head, the less is available for removal in the lungs. Thus, the extent of collec- tion in the upper respiratory tract affects deposition in more distal regions. Figure 2-3 depicts deposition in the tracheobronchial tree. The relation be- tween deposition and particle size is not as well defined as in other regions; fractional tracheobronchial deposition is relatively constant over a wide range of particle size. Deposition in the pulmonary region (al- veolated airways) is shown in Figure 2- 4. With oral inhalation, deposition in- creases with particle size after a mini- mum at approximately 0.5 Am. With nasal breathing, percent deposition tends to decrease with increasing particle size. The removal of particles in more proximal airways determines the shape of the pul- monary deposition curves. For example, increased upper respiratory and tracheo- bronchial deposition would be associated with a reduction of pulmonary deposition; thus, nasal breathing results in less pul- monary penetration of larger particles, and a smaller fraction of deposition of entering particles, than does oral inhala- tion. In the latter case, the peak for pul- monary deposition shifts upward to larger particles and is more pronounced. However, with nasal breathing, there is a relatively constant pulmonary deposition over a wider range of particle size. Rodents are often used in aerosol in- halation studies. To apply the results 60: 50: 40 - O 30 20t 10 . 21 to humans adequately, it is essential to consider interspecies differences in total and regional deposition patterns. In evaluating studies with aerosols, the amount of deposition expressed merely as a percentage of the total inhaled-i.e., deposition efficiency-might not be ade- quate information for relating results between species. For example, total respiratory tract deposition for par- ticles of the same size can be similar between humans and many laboratory ani- mals; total deposition efficiency is in- dependent of body size (McMahon et al., 1977; Brain and Mensah, 1983~. Different species exposed to identical particles at the same exposure concentration will not receive the same initial mass deposi- tion. If the total amount of deposition is divided by body weight, smaller animals would receive greater initial particle burdens per unit weight per unit exposure time than would larger ones. Humans differ from most other mammals used in inhalation pharmacologic studies in various aspects of respiratory tract anatomy; but the implications for par- ticle deposition have not been adequately understood. One major difference is in bronchial-tree branching pattern, which might affect the depth of penetration of inhaled particles, as well as localized patterns of deposition. In the pulmonary region, alveolar size differs between species; this could affect the probabili- ty of deposition by diffusion and sedimen ~Orai Inhalation _ 60 En o 0.01 0.1 1.0 10 PARTICLE DIAMETER (pm) FIGURE 2-3 Particle deposition efficiency in human tracheobronchial tree. Values are means with stand- ard deviations. Source: Reprinted with permission from Schlesinger, 1989. 40 30 20 10

OCR for page 17
22 70 60 50 to ~ 40 of o o CL , 3 _ 20 O 1 1 1 111111 1 1 1 111111 1 1 1 1 1111 O 0.01 0.1 1.0 10 PARTICLE DIAMETER (firm) ~ Oral Inhalation Nasal Inhalation _ 70 IT 60 in ~ AN AAA ~ 1 1 40 30 20 in ration, because of interspecies differ- ences in the distance between airborne particles and alveolar walls. From the discussion of deposition mech- anisms, it should be evident that the major particle characteristic that in- fluences deposition is size. A particle characteristic that can alter its size after inhalation is hygroscopicity. Hy- groscopic particles grow substantially while they are still airborne in the respi- ratory tract and are deposited according to their hydrated size, rather than their initial dry size. Some environmental particles consist of a relatively insoluble core coated with various chemical substances, such as metals, acids, and organic compounds. Variations in both the core material and any adsorbed material depend on the source of the particles. For example, in combus- tion processes, volatile metal compounds and organic compounds might condense on carbon particles during the cooling of the effluent stream in the smokestack or exhaust line and during release to the atmosphere. Adsorption or condensation of gases from the atmosphere can produce a high surface concentration on particles that are already airborne. If those proc- esses are diffusion-limited, the conden- sation and coagulation will be quantita- tively proportional to particle diameter MARKERS IN PULMONARY TOXICOLOGY FIGURE 2-4 Particle deposition efficiency in human alveolated anways. Values are means with standard d~iabons Source: Rep0ed web pi from Schlesinger, 1989. for particles with Dae larger than 0.5 ,um and proportional to particle surface area for smaller particles. In either case, the fractional mass of the surface-coating material will be greater on smaller particles than on larger ones. Thus, surface deposition provides a layer of soluble material present at high con- centration and results in small-particle enrichment, which leads to a shift in size of the potentially toxic surface materi- als to Dae smaller than the Dae of the total particle mass. Consequently, the deposi- tion of surface-enriched material could be significantly increased. Gases Because of the rapid and consequent even distribution motion of gas molecules in air, inhaled gases can be deposited on or at least come into contact with a large portion of the surface of the respiratory tract. However, if deposited molecules are not removed by metabolic action, solu- bility in or chemical reaction with the epithelial lining fluid, or absorption into blood, the molecules will be re-en- trained in the airflow and reach more dis- tal parts of the respiratory tract. Depo- sition efficiency of a gas is usually de- fined as the proportion that is retained by the respiratory tract. Gases that are

OCR for page 17
EXPOSURE water-soluble and highly reactive chemi- cally will be deposited (retained) to a higher degree in the upper respiratory tract, only at high concentrations will they reach the lung. Lipid-soluble, non- reactive gases, such as anesthetics, will be deposited in the alveoli, even at low concentrations. The deposition and absorption of in- haled lipid-soluble gases have been in- vestigated for some time by anesthetiolo- gists (Fiserova-Bergerova, 1983~. As noted above, lipid-soluble vapors are not absorbed by the aqueous lining fluid of the upper respiratory tract and so pene- trate to the alveolar region at low con- centrations. Their absorption from the respiratory tract depends mainly on their solubility in blood. Removal from the blood is by exhalation or metabolic ac- tion. A compound with low blood solubility and low tissue metabolic rate, such as vinylidene fluoride, will quickly reach saturation concentrations in the blood that depend on Henry's law and the gas- blood partition coefficient. The amount of such a compound that reaches other tis- sues depends on perfusion, or blood flow, because only a small amount will dissolve in blood. However, if a compound has high solubility in blood, such as acetone, the amount that reaches other tissues will depend on ventilation, because the blood will take up a large portion of any highly soluble chemical that enters the lung. A ventilation-perfusion model similar to that developed for alveolar deposition of anesthetic (Henderson and Haggard, 1943, pp. 71-89) has been reported for the deposition of acetone and ethanol (non- reactive vapors) in the upper respiratory tract of the rat and guinea pig (Morris and Cavanagh, 1986; Morris et al., 1986~. Species differences were observed: the rat was more efficient in deposition of the gases (Morris et al., 1986~. Deposi- tion of ethanol was twice as efficient as deposition of acetone, in agreement with the higher partition coefficient of etha- nol and in agreement with the results of earlier studies of ethanol deposition in humans (Landahl and Hermann, 1950~. Water-soluble, reactive gases, such as the rat carcinogen formaldehyde (Kerns 23 et al., 1983) are deposited mainly in the nasopharyngeal region. Less water-solu- ble reactive gases, such as ozone and ni- trogen dioxide, penetrate more deeply into the respiratory tract and damage the terminal bronchioles. Miller and coworkers (F. J. Miller et al., 1982, 1985) have done extensive quan- titative extrapolation modeling of the deposition of reactive gases in the respi- ratory tract. Their model takes into ac- count not only the physical and chemical properties of an inhaled gas, but also the anatomic and physical properties of the respiratory tract and their influence on the deposition and absorption of in- haled gases. The model predicts that dosimetry and uptake of ozone or other reactive gas will depend heavily on the thickness of the mucous blanket and on the rate of chemical reaction of the gas with the lining materi- al. The model also predicts that exercise (with its increased ventilation rate) will increase the dose proportion of a reactive gas that reaches the lung. CLEARANCE OF INHALED MATERIAL FROM THE RESPIRATORY TRACT Insoluble Particles The toxic response to inhaled particles depends on both the amount of material deposited at target sites and the duration of retention of deposited material. Par- ticles are cleared from their deposition sites by various routes and interacting processes. The specific pathway depends on the region of the respiratory tract where the material is deposited. The primary biologic mechanisms of clearance of insoluble particles are mu- cociliary transport in the nasal passages and tracheobronchial tree and removal from the pulmonary region by resident mac- rophages. Most of the surface of the tra- cheobronchial tree and nasal passages is lined with ciliated epithelium overlaid with mucus. The mucus is produced by spe- cialized epithelial cells and submucosal glands and consists of two layers: a low- viscosity hypophase that surrounds the

OCR for page 17
24 cilia and within which they move and a high- viscosity epiphase that lies on top of the cilia. Material deposited on the mucus is cleared to the pharynx by movement of the epiphase due to coordinated beating of the cilia. Several mechanisms and pathways con- tribute to clearance from the pulmonary region, but their relative importance is uncertain. The mechanisms involve ab- sorptive (dissolution) and nonabsorp- tive processes, which can occur simulta- neously or at different times. Nonabsorptive clearance processes are mediated primarily by alveolar macro- phages. These large mononuclear cells originate as precursors in bone marrow, reach the lung as monocytes, and mature in the pulmonary interstitium, from which they traverse the epithelium to reach the alveolar surface. As macrophages move freely on alveolar surfaces, they phago- cytose, transport, and detoxify deposited material, with which they come into con- tact by chance or by directed motion due to chemotactic factors. Particle-laden macrophages are cleared from the pulmonary region along a number of pathways. The primary route is the muco- ciliary system, but the mechanism by which cells reach it is not certain. One possi- bility is movement along the alveolar epi- thelium; another involves passage through the alveolar epithelial wall into the in- terstitium-macrophages could then reach the surface of ciliated airways, perhaps through small collections of lymphatic tissue at alveolobronchiolar junctions. Particle-laden macrophages that do not clear by way of the bronchial tree might actively migrate within the interstitium to a nearby lymphatic channel or, with uningested particles, be carried in the flow of interstitial fluid toward the lym- phatic system. Alternatively, unin- gested particles or macrophages in the interstitium could cross the alveolar capillary endothelium and enter the blood directly. Finally, free particles or mac- rophages within the interstitium could end up in perivenous or subpleural sites, where they become trapped. The migration and grouping of particles and macrophages can lead to the redistribution of deposits into focal aggregates. MARKERS IN PULMONARY TOMCOLOGY The most important mechanism of absorp- tive clearance is dissolution. Particles that dissolve in the alveolar fluid can diffuse through the epithelium and inter- stitium into the lymph or blood, and par- ticles initially translocated to and trapped in interstitial sites can undergo dissolution there. Dissolution is a major clearance route even for particles usual- ly considered to be relatively insoluble. The factors that affect the solubility of deposited particles are poorly under- stood, although they are influenced by the particles' surface-to-volume ratio and other surface properties. Some depos- ited material can dissolve after phago- cytic uptake by macrophages. For example, some metals can dissolve within the acidic . milieu of phagosomes. It is not certain, however, whether the dissolved material then leaves the cell. The residence time of deposited par- ticles depends on their clearance route. Material deposited on the conducting air- ways is cleared within about 1-2 days, although some long-term retention can occur. Particles deposited in the pul- monary region might remain for months or years or be retained indefinitely in in- terstitial sites. Soluble particles, and even particles with relatively low solubility, can dissolve in the pulmonary region. Solubilized components can be retained in the lungs, be redistributed in the body (where they might be retained in extrapulmonary tissues), or be ex- creted. In the conducting airways, solu- bilization occurs if the rate of dissolu- tion is greater than the rate of removal by mucus transport. The mucociliary system of the lung provides a major line of defense in eliminating bacteria, inhaled par- ticles, toxicants, and cellular debris. Bates ( 1989) has published an excellent review of the physiology of mucociliary clearance and its function in protecting the lung. The present subcommittee has addressed mucoculiary clearance in sever- al sections, depending on the type of contaminant and the the of marker being considered. - , c, The retention of some materials cannot be studied experimentally in humans, so experimental animals must be used. Dosim- etry depends on clearance rates and routes,

OCR for page 17
EXPOSURE so adequate pharmacologic assessment necessitates relating clearance kinetics in animals to those in humans. Although the basic mechanisms of respiratory tract clearance are similar in humans and most other mammals, regional clearance rates show substantial variation among species, even for similar particles deposited under comparable exposure conditions (Snipes et al., 1983~. Dissolution rates and rates of transfer of dissolved substances into blood are probably related solely to the properties of the material being cleared and essentially independent of species (Cuddihy et al., 1979; Griffith et al., 1983; Bailey et al., 1985~. However, dif- ferent rates of mechanical transport, such as macrophage clearance from the pulmonary region (Bailey et al., 1985) and mucocili- ary transport in conducting airways (Felicetti et al., 1981), occur and result in species-dependent rate constants for these clearance pathways. Differences in regional (and perhaps total) clearance rates among species are probably due to the latter processes. Gases and Soluble Particles Reactive gases-such as ozone, nitrogen dioxide, and formaldehyde-exert their toxic effects at the site of deposition in the respiratory tract. Little is known about distribution of these gases or their products beyond the site of deposi- tion. Therefore, the discussion of the clearance or disposition of inhaled ma- terials in this section will include only nonreactive gases and soluble particles. The collection of knowledge about the extent and rate at which inhaled toxic materials distribute throughout the body and are excreted is referred to as toxico- kinetics. Toxicokinetic studies are fun- damental to an understanding of internal dose and dose to target tissue. Toxicokin- etic measurements are specific for indi- vidual pollutants and thus are valuable as biologic markers of environmental exposure. For ethical and practical reasons, most detailed toxicokinetic studies have been performed in animals. In such studies, 25 either radiolabeled material can be used to detect and measure the parent substance and its metabolites or standard analytic chemistry techniques can be used to meas- ure the same materials. Using newer forms of mathematical modeling, which include physiologic parameters, one can make reas- onable extrapolations from animal data to humans. This section discusses the types of toxicokinetic data that can be obtained in animal studies. The follow- ing sections discuss the types of human samples that can be analyzed to obtain information on exposure history, internal dose, and dose to target tissue and how animal toxicokinetic data can be extrapo- lated with modeling techniques to predic- tions for humans. In animal inhalation exposure studies, one can determine the fraction of an in- haled substance that is absorbed, the time it takes to reach a steady-state con- centration of the substance and its metab- olites in the blood, equilibrium concen- trations in tissues, major routes and rates of excretion of the substance and its metabolites, and times required for their elimination from each tissue and from the whole body. One can also deter- mine the effects of exposure concentra- tion, of exposure rate, and of repeated exposures on those measures. Tissue and excrete samples can be analyzed for mater- ials of interest with standard analytic chemistry techniques or, for greater sen- sitivity, with radiolabeled compounds. In the latter case, the chemical form of a labeled compound (exposure material or metabolite) is often identified. A few examples will illustrate the im- portance of such data in determining the internal dose of a compound received by an organism and the dose to target tis- sue. In rats exposed to methyl bromide at atmospheric concentrations of 50, 300, 5,700, and 10,400 nmol/L, the internal doses of the compound at the two highest exposure concentrations were equal (Medinsky et al., 1985~. That was because the absorbance of methyl bromide and the tidal volume were decreased at the highest exposure concentration. The data indi- cate that absorption of methyl bromide is a saturable process and show the impor

OCR for page 17
26 lance of knowing both internal dose and external exposure concentration. When rats and mice were exposed by in- halation to formaldehyde at 14.3 ppm for up to 2 years, the rats had a 50% incidence of nasal carcinoma, the mice an incidence of only 1% (Kerns et al., 1983~. The dif- ference was explained biologically by analysis of effective dose, as opposed to administered dose, the external expo- sure concentration (Starr and Gibson, 1984~. The mice were more sensitive to the sensory irritation properties of formal- dehyde than the rats and thus had a smaller minute volume during exposure and received a lower internal dose (Barrow et al., 1983~. Such species differences can often be explained by toxicokinetic data, particu- larly if rates of formation and elimina- tion of metabolites are determined. Studies in rats and mice exposed to benzene indicated that the mice had higher tissue and blood concentrations of putative tox- ic metabolites of benzene than did rats (Sabourin et al., 1987a,b). Mice were also more sensitive to benzene in long- term bioassay studies (NTP, 1986) and to the tumorigenic properties of inhaled butadiene (Huff et al., 1985~. Pharmaco- kinetic studies on butadiene and its meth- ylated derivative, isoprene, indicated higher blood concentrations of the reac- tive epoxide metabolites in mice (the more sensitive species) than in rats exposed at the same atmospheric concentration (Bond et al., 1986; Dahl et al., 1987~. It should be borne in mind that the lung is a target organ for some toxicants that can reach the lung through the blood or skin. For example, prolonged skin expo- sure, inhalation, or ingestion of the herb- icide paraquat can cause death from lung injury in humans and animals. Paraquat accumulates selectively in lung tissue by a carrier-mediated mechanism and is retained there; accumulation not only influences lung paraquat burden, but is probably also an important determinant of organ response. Other pneumotoxic agents can reach the lung through the cir- culation, including antibiotics (e.g., Neomycin) and plant toxins (e.g., elec- trophilic metabolites of pyrrolizidine M'4R=RS IN PULMONARY TOXICOLOGY alkaloids). Therefore, in the considera- tion of pulmonary markers and their devel- opment, it is important to examine not only inhaled environmental toxicants, but also those which reach the lung through the blood, through the skin, or by ingestion. MONITORING FOR INHALED MATERIAL Several biologic samples can be ob- tained from humans to assess internal dose or dose to target tissue. In review- ing the biologic approaches to dosimetry of carcinogens in humans, Tannenbaum and Skipper ~ 1984) listed blood, urine, feces, sweat, hair, nails, milk, semen, saliva, lens, and biopsy tissues. Respira- tory system samples-such as exhaled air, nasal-ravage fluid, and, in special cases, bronchoalveolar- lavage fluid- could be added. Substances most suitable for field sampling in epidemiologic studies are blood, urine, hair, nails, saliva, exhaled air, and perhaps nasal- lavage fluid. The other substances are more likely to be sampled in laboratory studies. Biologic monitoring of industrial work- ers is most often based on blood, urine, or exhaled air (Lauwreys, 1983~. Such an approach is appropriate in an industri- al setting, because samples can be taken often and the exposures are normally high- er than in an environmental setting. The methods also provide information for those in environmental research on the relationship between magnitude of expo- sure and the amount of compound or metabo- lite expected to appear in body fluids. However, for environmental exposures, such analyses might not be sensitive enough to detect small exposures; for com- pounds cleared rapidly from the body, only the most recent exposures can be de- tected. Some newer methods, however, have proved useful in monitoring for chemical exposures. The following sections discuss the types of monitoring that can be done with such samples from humans and the kinds of animal studies on which some human moni- toring is based. The emphasis is on newer techniques and on samples that are most

OCR for page 17
EXPOSURE relevant to exposure by environmental inhalation. Insoluble Particles The best marker of exposure to particles that one could hope for is detection of the inhaled material at the sites in the lung where disease develops. Establish- ing the presence of the particles at such sites is good; quantitation is better, if it is important to determine the dose delivered to a target site. In rats and mice exposed to asbestos, fiberglass, wollastonite, iron, silica, and ash from the volcano Mt. St. Helens, it has been established that 80% of the particles small enough to pass through the conduct- ing airways was deposited on the bifurca- tions of alveolar ducts (Brody and Roe, 1983~. Scanning electron microscopy was used after brief exposures (1, 3, or 5 hours), to calculate the number of par- ticles per square micrometer of bifurca- tion surface (Brody and Roe, 1983~. That number is a marker of exposure. Whether such a marker could be useful for human exposures is difficult to know. The lung would have to be fixed for electron micros- copy within several hours after exposure; otherwise, substantial numbers of inhaled particles would have been transported from the alveolar surfaces by epithelial cells, macrophages, and the alveolar lin- ing layer (Brody et al., 1981~. Particle deposition is a good marker of exposure in animals, because it can predict whether the subject is likely to develop lung disease, where the disease will originate, and the nature of the path- ogenic process. The first prediction is based on the elemental nature of the inhaled particles as they reside on the epithelial surfaces. That is determined routinely with x-ray spectrometry, a technique widely used in studies of par- ticle burden in humans and animals (Brody, 1984~. If the particles are asbestos fibers or silica crystals, it would be valid to assume that a fibrogenic dis- ease will ensue. If wollastonite fibers or ash particles are detected, it is less likely that a pathologic response will follow. The second prediction is validat 27 ed by a series of studies that show that the initial response of epithelial cells, macrophages, and fibroblasts takes place at the sites of original particle deposi- tion, i.e., the alveolar duct bifurca- tions (Warheit et al., 1986~. Decades ago, pathologists used the finding of early inflammation and fibrogenesis in the bronchiolar-alveolar regions as a marker of particle-induced lung disease (Wagner, 1965~. The nature of the response should be predictable on the basis of the two other predictions. If asbestos is inhaled, then interstitial macrophage- mediated fibrogenesis should be expected. If silica is inhaled, a nodular fibrosis mediated by acute and chronic inflamma- tory cells will develop. If iron or ash particles reach the alveolar surfaces, one should expect rapid clearance of par- ticles with little or no pathogenic sequelae. Some particles are retained in the lungs for long periods, even through the life- time of the experimental animal or occupa- tionally exposed person (Abraham, 1978~. Such particles are excellent markers of exposure and, as suggested above, could increase understanding of the nature of any disease process that is present. For example, if rats inhale chrysotile or cro- cidolite asbestos for 1 hour, approxi- mately 20% of it will still be in the lungs a month after exposure (Roggli and Brody, 1984; Roggli et al., 1987~. If clearance continues as predicted by calculated clearance curves (Lippman et al., 1980; Roggli and Brody, 1984), the animals will still have many fibers in their lungs at the time of expected natural death. Oc- cupationally exposed people have large quantities of dust in their lungs many decades after cessation of exposure (Selikoff and Hammond, 1978~; this know- ledge has yet to be exploited in a quantita- tive way in attempts to use lung burden to predict lung injury. Many techniques are available for as- sessing the lung burden of many commonly inhaled particles. Several new imaging techniques can be used to determine the nature of crystalline particles. Transmission electron microscopy (TEM) is used to locate the particles in lung

OCR for page 17
32 be monitored for markers of exposure or effects include sputum, nasal-ravage fluid, and bronchoalveolar-lavage (BAL) fluid. Sputum and nasal-ravage fluid can be obtained with relatively noninvas- ive procedures. BAL enables sampling of alveolar lining fluid, which is in direct contact with or intimately related to cells that are involved in injury or dis- ease. However, the procedure is relative- ly invasive, requiring that a subject un- dergo fiberoptic bronchoscopy with some anesthetic. Although the procedure is relatively safe in normal or asymptomatic subjects, substantial morbidity might occur in a person with severe pulmonary disease. Asbestos and wool fibers have been found in sputum from exposed people. In fact, the ability to detect asbestos bodies in BAL fluid from those exposed occupational- ly to asbestos is being used diagnostically (De Vuyst et al., 1987~. Other inhaled particles, such as coal dust or mineral dust, should be detectable in sputum, nas- al-lavage fluid, and BAL fluid from heavily exposed persons (Roggli et al., 1986~. The sensitivity of such monitoring and the relationship between what is in the samples and the extent of exposure are unknown. Recent studies have shown BAL to be a sensitive indicator of effects of envi- ronmental pollutants on the lung. Koren and co-workers (1989) reported that, 18 hours after a 2-hour exposure to ozone at 0.4 ppm, the proportion of polymorpho- nuclear leukocytes in BAL fluid from healthy, nonsmoking men increased by a factor of 8. Similar but smaller increases were seen in immunoreactive neutrophil elastase. Markers of vascular permeabil- ity in the BAL fluid doubled. Complement fragment C3a increased by a factor of 1.7, prostaglandin E2 by a factor of 2, fibro- nectin by a factor of 6.4, and urokinase plasminogen activator by a factor of 3.6. Smaller exposures (at 0.1 ppm for 7 hours) produced smaller responses, but there were no definite indications of a thresh- old (above ambient concentration) for ozone-related tissue responses. The use of such samples as quantitative monitors of exposure will require valida AL4RKERS IN PULMONARY TOXICOLOaY tion in humans under conditions of known exposure. With the availability of more sensitive means of detecting specific adducts, it might be possible to monitor human samples previously thought to con- tain nondetectable concentrations of exposure-related material. Pulmonary macrophages represent the cleanup crew of the lung and lower airways and can be expected to reflect the material inhaled and deposited in that area. Sputum, bron- chial washes, and BAL fluid contain macro- phages. Macrophages are not target sites for tumorigenic responses, but might be used to monitor the extent of recent ex- posures. Exposures to insoluble particles or fibers should be reflected as phagocy- tosed material in the macrophages. Par- ticle-associated organic substances are retained in the lung long enough to be me- tabolized (Sun et al., 1983, 1984~. Bond et al. (1984) showed that macrophages can metabolize such compounds as benzo~a]- pyrene to reactive substances that could bind to DNA. Newer methods, such as those used by Haugen et al. (1986) to detect ben- zoLa~pyrene adducts in lymphocytes, could allow detection of DNA adducts in alveolar macrophages. The average time that a mac- rophage spends in the alveolar space has been estimated at 7-27 days (Van oud Alblas and van Furth, 1979; Bowden, 1983~. Thus, alveolar macrophages could be used to meas- ure cumulative exposures over a relative short period (days to weeks). Such an ap- proach deserves further investigation. Exhaled Air Exhaled air contains an array of vola- tile organic constituents that are likely to be in equilibrium with a number of com- partments in the lung or can arise from endogenous or absorbed volatile sub- stances circulating in the blood. In addi- tion, some substances in lung air might be in equilibrium with alveolar lining material. Finally, cells within the air- spaces (including mucous glands) and cells that are attached to the bronchial epithelium (such as alveolar macrophages) could also contribute to the constituents of lung air. Obviously, exhaled air lends itself to easy noninvasive collection.

OCR for page 17
EXPOSURE Exhaled air can be analyzed with gas chromatography/mass spectroscopy (GC/MS) to yield markers of exposure in the form of volatile substances in the blood. In one case, differences in the contents of exhaled air of a nonsmoking submariner before and after a cruise gave information on the environment of the submarine (Knight et al., 1984, 1985~. The halogen- ated hydrocarbons used for refrigeration on the submarine were easily detectable in the submariner's breath. The exhaled air contained both endogenous metabolites and atmospheric contaminants from the submarine. The comparison between the before-cruise sample and the after-cruise sample helped to distinguish between the metabolites and contaminants. For exam- ple, isoprene, the monomeric unit of ter- penes known to be an endogenous metabolite in mammals and known to be emitted by a wide range of plants (Tingley et al., 1979), was present in the exhaled breath both before and after the cruise. Conkle et al. (1975) reported the trace contamin- ants in exhaled air from eight unexposed volunteers. They identified 53 volatile compounds and used a cryogenic trapping system for concentrating trace organic compounds to allow detection of submicro- gram quantities. Such a procedure appears to have the sensitivity required for ap- plication to environmental exposures. In the Total Exposure Assessment Method- ology (TEAM) study funded by the Environ- mental Protection Agency (EPA) (Wallace, 1987), the amounts of 11 prevalent vola- tile organic compounds found in the breath of 355 New Jersey residents were found to correlate with the previous 12- hour average air exposures. That caused the investigators to conclude that"breath measurements may be capable of providing rough estimates of preceding exposures." The same group used analysis of exhaled air to determine exposure to benzene during the filling of a gasoline tank, exposure to tetrachloroethylene in dry-cleaning shops, exposure to chloroform from hot water in the home, and exposure to aromatic compounds in tobacco smoke. A potential approach that has been lit- tle studied is the analysis of exhaled air for volatile metabolites that might 33 be involved in lung disease. The approach is noninvasive and involves sampling of volatile organic compounds from easily obtainable physiologic materials, such as breath and saliva. There are two criti- cal requirements for this type of analysis to be useful: the disease process must lead to the production of volatile metabo- lites that will be present in exhaled air, and these metabolites must reach measur- able concentrations in the total exhaled air. Fulfillment of the latter require- ment is limited by the sensitivity and sophistication of the instruments used to analyze the exhaled air. With the use, for example, of gas chromatography com- bined with mass spectroscopy and appro- priate computer analysis, the detectable amount could be as small as several hundred picograms. Many volatile constituents of body fluids have been characterized in dia- betes, respiratory viral infections, and renal insufficiency (Zlatkis et al., 1981~. In several diseases, breath analy- sis with GC/MS has revealed the presence of simple endogenous alcohols, ketones, and amines and numerous compounds of endog- enous origin (Chen et al., 1 970a,b; Krotoszynski et al., 1977; Simenhoff et al., 1977; KaJi et al., 1978~. For exam- ple, concentrations of mercaptans and C2-C5 aliphatic acids are increased in the breath of patients with cirrhosis of the liver (Chen et al., 1 970a,b; Kaji et al., 1978), and dimethyl and trimethyl amines are present in the breath of uremic patients (Simenhoff et al., 1977~. A simi- lar approach to the analysis of other vola- tile metabolites involved in acute and chronic damage to the lung should be explored. Blood Inhaled organic compounds enter the blood directly from the lung, and analysis of the blood for an inhaled compound or its metabolites can provide valuable in- formation on recent exposures. Tradi- tional analytic chemistry techniques involving various types of chromatography and spectroscopy have been used to sepa- rate and identify compounds. Recently,

OCR for page 17
34 an innovative vacuum-line distillation technique has been used to separate vola- tile compounds and their volatile metabo- lites in blood (Dahl et al., 1984~. The approach has proved useful for toxicokin- etic studies of such compounds as buta- diene (Bond et al., 1986) and isoprene (Dahl et al., 1987-compounds that have volatile monoepoxide and diepoxide metabolites. To be able to monitor cumulative expo- sures over longer periods, one needs a marker that is not cleared from the blood as rapidly as are organic compounds and their metabolites. Newer approaches to biologic monitoring make use of the fact that reactive metabolites of organic com- pounds can react spontaneously with nucle- ophilic sites on macromolecules to form covalently bound adducts to DNA, hemo- globin, and other important proteins. Once formed, the adducts are relatively long-lived in the body (compared with the exposure compound or its free metabo- lites). Some of the materials of greatest interest, mutagens and carcinogens, are classes of compounds that either are reac- tive electrophiles or can be metabolized to electrophiles. The electrophiles can then bind to nucleophilic macromolecules, such as proteins and DNA. Blood contains large amounts of two proteins, albumin and hemoglobin, with reactive amino and sulfhydryl groups that can interact with electrophiles. Adducts formed with such proteins can be expected to remain in the blood with the same half-time as the pro- teins. In the case of hemoglobin, the life span of the protein, approximately 4 months, permits detection of cumulative doses from exposures over several months. Albumin has a shorter half-life-approxi- mately 2 weeks. Recent research (Harris et al., 1987; Poirier and Beland, l 987b) has centered around the use of hemoglobin adducts for monitoring. Hemoglobin adducts formed from reactive metabolites are being investigated for their potential as biologic markers of exposure. Segerback (1983) reported the formation of hemoglobin adducts in mice exposed to ethene or ethylene oxide. The same group has published many studies on the ability of alkylating agents to form MARKERS IN PULMONARY TOXICOLOGY hemoglobin adducts and on the value of monitoring persons exposed to such agents by quantification of these adducts (Osterman-Golkar et al., 1976, 1983; Calleman et al., 1978~. Tornqvist et al. (1986) used hemoglobin adducts to deter- mine tissue doses of ethylene oxide in cigarette-smokers. They found an in- creased amount of hydroxyethylation of the N-terminal valine of hemoglobin that correlated with the amount of ethene in cigarette smoke. Pereira and Chang (1981) studied a series of 15 carcinogens and their ability to form hemoglobin ad- ducts in animals. They found that oral exposures of all the carcinogens produced hemoglobin adducts and that the most effi- cient binding occurred at the lowest doses (0.1 ~mol/kg). Green et al. (1984) found the use of hemoglobin adducts prom- ising for monitoring arylamine exposures in humans, on the basis of a study in rats exposed to 4-aminobiphenyl. Shugart and Matsunami (1985) found that hemoglobin adduct formation provided a suitable mon- itor of exposure to benzo~a~pyrene in mice. Newmann (1984), in a review article, pointed out that hemoglobin adduct forma- tion not only is a more reliable indicator of internal dose, but also indicates a person's capacity to metabolize an in- haled organic compound to a reactive in- termediate. To judge from the results of studies conducted thus far, the use of hemoglobin adducts to monitor exposure to chemicals is promising. The adducts form with a vari- ety of compounds and allow detection of subnanogram amounts of material. The for- mation of albumin adducts should also be investigated as a potential marker of ex- posure. Despite the short half-life of albumin, this protein is available in the serum for reaction with reactive me- tabolites without the need for the metabo- lites to cross a cellular membrane. In occupational settings, where small ex- posures can occur daily, an end-of-shift measurement of albumin adducts would yield a sensitive marker of exposure. Urine Recent findings have suggested that

OCR for page 17
EXPOSURE biochemical markers in urine can be useful in monitoring the development of nonneo- plastic pulmonary diseases. Clearly, such markers would be advantageous if they could be obtained with noninvasive proce- dures. However, their major drawback is that they reflect events in the lung only indirectly. Microsomal metabolism of inhaled or- ganic compounds can produce water-soluble metabolites and their conjugates that are excreted in the urine. These compounds are not usually the toxic forms of an in- haled chemical-although exceptions occur (see Chellman et al., 1 986-but their presence in urine can indicate that expo- sure to a specific chemical has taken place. If pharmacokinetic information is available from animal studies and from physiologic modeling, the total amount of the metabolites excreted in the urine can be used for quantitative estimates of exposure. For example, phenol in urine has long been used to monitor worker ex- posure to benzene (Teisinger et al., 1955~. More recently, DNA adducts have been monitored in urine (Groopman et al., 1985) with monoclonal antibodies to ad- ducts of aflatoxin Be. Urine has also been analyzed for muta- genic activity as a monitor of exposure to genotoxic agents (Bloom and Paul, 1981~. The relationship between concen- trations of urinary mutagens and risk of cancer is still unknown and requires fur- ther research. Analysis of urinary muta- gens shows promise, on the basis of such work as that of Camus et al. (1984) in which two strains of mice with different suscep- tibilities to development of cancer were treated with benzo~alpyrene and the uri- nary mutagen concentrations were cor- related with tumor formation. In general, the ability to produce urinary mutagens corresponded to susceptibility to tumorigenesis. The use of urinary mutagens to monitor environmental exposure will be difficult, because such confounding factors as diet, smoking, and occupational exposure intro- duce uncertainties in interpretation of the assays. Ohyama et al. (1987) reported that ingestion of cooked salmon increased urinary mutagens, whereas ingestion of 35 cooked vegetables did not. Kawano et al. ( 1987) reported a correlation between increases in mutagens in smokers' urine and the number of cigarettes smoked per day. Similar results were reported by Mohtashamipur et al. (1985~. T. H. Conner et al. (1985) reported no increase in uri- nary mutagens in autopsy-service workers exposed to formaldehyde, but there was an increase in a heavy smoker in the con- trol group. Steel workers exposed to coke- oven emission (benzo~a~pyrene concentra- tions, 0.01-0.6 ~g/m3) had slightly higher concentrations of urinary mutagens than unexposed controls, but smoking habits were the major influence on the concentra- tions (De Meo et al., 1987~. Turnover of extracellular matrix of the lung has been suggested as accompany- ing tissue remodeling after exposure to environmental pollutants. Some inves- tigators believe that measurement of hydroxyproline or hydroxylysine in urine could reflect collagen turnover; both these amino acids are found only rarely in molecules other than collagen. Kelly et al. (1986) tested the utility of connec- tive tissue breakdown products as markers of current injury after brief exposure of Fischer rats to NO2. They found a linear increase in hydroxylysine excretion with increasing NO2 concentration. Several environmental contaminants can now be monitored in urine. For example, Enterline et al. (1987) have examined exposure to arsenic in men working at a copper smelter in Tacoma, Washington. They found a good correlation between uri- nary concentrations of arsenic in workers (as a marker of exposure) and respiratory tract cancer. In clinical studies by Hatton et al. (1977), hydroxyproline glycosides were increased in the urine of three American astronauts who were accidentally exposed to toxic fumes of NO2 during the descent phase of their mission. More recently, Yanagisawa et al. (1986) attempted to cor- relate NO2 exposure with urinary hydroxy- proline-to-creatinine ratios in 800 women who were mothers of primary-school children in two communities around Tokyo. The ratio was found to be correlated with NO2 exposure and with numbers of cigarettes

OCR for page 17
36 smoked actively and passively. Such ex- periments appear promising, but methods must be developed to allow identification of specific sources (e.g., lung, liver, and bone) of the breakdown products, inas- much as collagen is present in many organs other than the lung. Studies to define site specificity of the breakdown products are essential. Adipose Tissue Organic vapors and gases that are in- haled can be retained in the fat depots of the body long after they have cleared from other parts of the body. Biopsies of fat could potentially be used to obtain information on lipid-soluble compounds to which a person has been exposed. An attempt to use that approach to de- termine which Vietnam veterans had been exposed to Agent Orange was based on labor- atory animal studies that showed that the toxic contaminant of Agent Orange, 2,3,7, 8 - tetrachlorodibenzo - p- dioxin (TCDD), accumulates in fat. Four groups of male volunteers were included: five "heavily exposed" veterans, 20 men who believed that they had been exposed to Agent Orange in Vietnam, 11 who had not been in Vietnam and had had no other con- tact with Agent Orange, and three Air Force officers who had definitely worked with either Agent Orange or TCDD (Gross et al., 1984~. Samples of 10-30 g of adi- pose tissue were taken from the abdominal wall of each volunteer and analyzed with gas chromatography and high-resolution mass spectrometry. The results indicated that it was possible to detect TCDD in human fat, but it was present in the fat of both the control (supposedly nonex- posed) men and the veterans. Four of the five heavily exposed veterans had a mean concentration of TCDD in fat of 55+34 parts per trillion (pot), and one did not have detectable TCDD. The mean in the other Vietnam veterans was 4 + 1 pot, and the mean in the control subjects was 6+ 3 pot. The 2,3,7,8-tetrachlorodibenzo-p- dioxin concentrations in the adipose tis- sue of Missouri residents distinguished between persons with a history of exposure MARKERS IN PULMONARY TOXICOLOGY to the chemical and control (presumable nonexposed) persons (Patterson et al., 1986~. Anderson (1985) stressed the im- portance of analyzing blood samples and adipose tissue samples at the same time to provide more information on the par- titioning between the two compartments. When sufficient information of this type is available, blood samples can be used to predict concentrations of the compound in fat. Some success has been achieved in moni- toring adipose tissues for evidence of exposure to polychlorinated biphenyl (PCB) congeners and for exposure to diox- ins and furans in accident situations, in which concentrations are higher than would be expected in ordinary environmen- tal exposures. M. S. Wolff et al. (1982) reported concentrations of PCBs in plas- ma and adipose tissue that were related to duration and magnitude of exposure in persons occupationally exposed to PCBs. Schecter et al. (1985) examined persons 1-2 years after exposure to PCBs, dioxins, and furans in the Binghamton, N.Y., State Office Building incident and found PCBs in their blood. The National Adipose Tissue Survey of the Environmental Protection Agency's National Human Monitoring Program (Lucas et al., 1982) is an example of the excel- lent use that can be made of tissue banks for retrospective studies of the influ- ence of occupation, geographic location, age, and sex on the concentrations of halo- genated hydrocarbons in human fat. DNA and Protein Adducts The use of DNA and protein adducts as a measure of exposure and of risk of tumor formation is under intense investigation (Poirier and Beland, 1987a). A rapid, sensitive method for detecting adducts has aided research (Randerath et al., 1985), but does not distinguish between types of adducts. The development of mono- clonal antibodies to specific DNA adducts promises to provide valuable monitoring tools for the future. Monoclonal anti- bodies have been developed for such speci- fic DNA adducts as those formed from ben- zofa~pyrene dial epoxide, 1-aminopyrine,

OCR for page 17
EXPOSURE 8-methoxypsoralen (Santella et al., 1987), and the carcinogen aflatoxin Be (Groopman et al., 1987~. Some of the most extensive work has been done on adducts formed after exposure to aflatoxin B1 (Groopman et al., 1987~. Studies have shown that the concentration of DNA adducts formed is quantitatively related to the intake of the carcinogen. In addition, the kinetics of the removal of the adducts have been determined, and a monoclonal antibody to the adducts has been made and can be used to measure them in DNA and in urine. A constant proportion of DNA adducts appears in urine as a func- tion of time after exposure (Groopman et al., 1985~. Such work to determine quan- titative relationships between exposure and adduct formation is needed for other carcinogens in the environment. Recent advances in analytic techniques allow the detection of DNA adducts in lym- phocytes. Haugen et al. (1986) reported the determination of polycyclic aromatic hydrocarbons (PAHs) in urine, benzoLa~py- rene dial epoxide-DNA (BPDE-DNA) adducts in lymphocyte DNA, and antibodies to the adducts in serum of coke-oven workers. To measure the adducts in lymphocytes, an ultrasensitive enzymatic radioimmuno- assay and synchronous fluorescence spec- trophotometry were used. Approximately one-third of the workers had detectable BPDE-DNA adducts in their lymphocytes and antibodies to epitopes on BPDE-DNA adducts in their serum. The concentrations of PAHs in the atmosphere of the workplace were 212-315,llg/m3. The problems and promise of the use of macromolecular adducts (both protein and DNA adducts) in biologic monitoring were reviewed at a recent workshop (Poirier and Beland, 1987b). The long-term hope is to be able to relate the concentration of adducts to the extent of exposure and to the risk of tumor formation. To achieve that, animal experimentation must indi- cate the relationship of adducts to the extent of exposure and tumorigenesis in various exposure regimens, and then the relationship for humans must be validated by adduct analysis of human tissues under known exposure conditions. In other words, the pharmacokinetics of adduct formation 37 must be known. Work addressing the rate of accumulation and persistence of DNA adducts, such as that reported by Belinsky and Anderson (1987) on the accumulation of 4-(N-methyl-N-nitrosamino)- 1 -~3- pyridyl)- 1 -butanone (NNK), is required. A linear relationship between exposure dose and adduct formation (with both pro- tein and DNA) has been shown in single- exposure experiments for hemoglobin adducts to 4-aminobiphenyl (Green et al., 1984), to alkylating agents (Osterman- Golkar et al., 1976, 1983), and to 4-di- methylaminostilbene (Newmann, 1984) and for DNA adducts to benzofaipyrene and to aflatoxin B: (Pereira et al., 1979; Appleton et al., 1982; Dunn, 1983; Adriaenssens et al., 1983~. With continu- ous exposures to alkylating agents (Swenberg et al., 1986), adduct concen- trations increase with time and eventual- ly reach a plateau or equilibrium where the rate of formation equals the rate of removal. Poirier et al. (1987) reported informa- tion that can aid in validating adduct formation in humans. DNA adducts with the cancer chemotherapeutic agent cispla- tin were measured in the nucleated peri- pheral blood cells and other tissues of patients receiving the drug. A positive correlation was observed between adduct concentrations in the blood cells and cumu- lative dose of the drug over several months. A major need that remains is to irlentifv the adducts that have biologic significance. Further research in animal models or in human clinical work, such as that reported by Poirier et al. ~ 1987) in cancer patients on chemotherapeutic regimens, is required to correlate adduct formation with biologic effects. DNA adducts have been detected in respi- ratory tract tissues, and the amount of these adducts increases after inhalation exposure to some organic compounds. DNA adducts identified as BPDE deoxyguanosine adducts were detected in the lungs of rats exposed to BaP (Wolff et al., 1989~. In another study, the concentration of total DNA adducts in various regions along the respiratory tract was compared with the site of tumor development in rats exposed to diesel exhaust. The concentration of

OCR for page 17
38 DNA adducts, detected with the 32P-post- labeling technique, was highest in the peripheral lung tissue, the site of tumors in rats chronically exposed to the exhaust (Bond et al., 1988~. Such results point to the importance of DNA adducts as meas- ures of effective dose of inhaled carcino- gens. However, it will be important to understand the kinetics of formation and repair of individual adducts (as opposed to total adducts), to elucidate the rela- tionship of DNA-adduct formation to carcinogenesis. Recent work has shown that adducts form- ed with protamine, a basic protein associ- ated with sperm DNA, might also be poten- tial markers of exposure at the critical site for adverse biologic effects. In one of the first reports linking adduct formation with a specific deleterious effect, Sega et al. (in press) identified one of the protamine adducts formed after acrylamide exposure as S-carboxyethyl- cysteine. The formation of that adduct acts to break S-S bonds in the protamine and might be the basis for the chromosomal breakage induced by acrylamide. MATHEMATICAL MODELING OF EXPOSURE Detailed toxicokinetic studies in hu- mans are usually impossible, because of ethical constraints and because not all relevant human organs and tissues can be sampled. However, mathematical modeling of the disposition and fate of inhaled chemicals in animals is useful for extra- polating data between species (NRC, 1987~. A strong impetus for the development of mathematical models of the deposition and retention of inhaled particles came with the nuclear age. The Task Group on Lung Dynamics of the International Commission on Radiological Protection developed such a model to determine the dosimetry associ- ated with inhaled radioactive particles (Task Group on Lung Dynamics, 1966~. That model, which remains the basis for model- ing the toxicokinetics of inhaled insolu- ble particles today, predicts the dose to internal tissues throughout the respir- atory tract and allows extrapolation of results from animals to humans. M4RKERS IN PULMONARY TOXICOLOGY Mathematical modeling has also been used successfully to predict the uptake, distribution, and elimination of inhaled lipid-soluble volatile materials (Andersen, 1981; Fiserova-Bergerova, 1983; Andersen and Ramsey, 1983~. Such models are useful for calculating doses to critical tissues, for extrapolating between species (including humans), for assessing hazards (F. J. Miller et al., 1987), and for guiding research. Recent advances in modeling that include physio- logic characteristics-such as blood flow into and out of an organ, membrane perme- ability, and chemical partitioning among blood, other tissues, and air-allow rather accurate extrapolations between species (Fiserova-Bergerova and Holaday, 1979; Fiserova-Bergerova et al., 1980; Andersen, 1981; Fiserova-Bergerova, 1983~. A pharmacokinetic model developed and validated in animals can be adjusted for the physiologic characteristics ap- propriate for humans and validated by an- alyzing readily available human materi- al, such as blood or excrete from humans exposed in clinical studies or in other situations of known exposure (Reitz, 1986). An example of that approach is the work of Ramsey end Andersen (1984), who devel- oped a physiologically based pharmacokin- etic model of the behavior of inhaled sty- rene in rats to predict accurately the behavior of inhaled styrene in humans. Using a set of physiologic and biochemical constants, the investigators were able to simulate the behavior of inhaled sty- rene in rats. Experimentally determined values for several measures were used: body weight, alveolar ventilation, blood flow rates, tissue volumes, blood-air partition coefficient, tissue-blood par- tition coefficients, maximal reaction rate (Vm:,,,j, and the Michaelis constant (substrate concentration at half the maxi- mal reaction rate). They then extrap- olated the values to humans and found that the model accurately predicted the amount of styrene that had previously been pub- lished to be in blood and exhaled air of humans exposed in clinical studies (Ramsey et al., 1980~. The same group (An- dersen et al., 1987) did a similar study

OCR for page 17
EXPOSURE later with methylene chloride and for the first time extended the model to include values to take into account the metabolic capacity of the lung, as well as the liver. This type of modeling constitutes a power- ful tool for extrapolating from animal to human data. Some precautions should be mentioned. In the animal studies, it is important to determine how and at what rate a chemi- cal and its metabolites are cleared from the body in different doses or exposure regimens, to find the range of doses over which the disposition and metabolism of the chemical are linearly related to dose. Such information is required for extrapo- lation from animal studies (normally at high doses) to the low doses normally en- countered by humans. It is also important to determine the effect of repeated expo- sure on the fate of a chemical. Repeated exposure at low concentrations, the most commonly encountered human exposure regi- men, could induce enzymatic changes that affect the toxicity of a chemical. Such studies can be conducted in animals. The biggest problem in using animal toxi- cokinetic data for human risk assessment is the potential for species differences in metabolism. Recent work by Medinsky et al. (in press a,b) extended physiologic modeling to include the disposition of both the parent substance and its metabo- lites in rats and mice given benzene orally or by inhalation. To extend such models to humans, however, requires some informa- tion on human metabolism of the chemical of interest. For example, rats and mice metabolize benzene differently (Sabourin et al., 1988~. Extension of the model of Medinsky et al. to humans by changing the physiologic parameters from those of the rat and mouse to those of humans necessi- tates choosing the appropriate metabolic parameters for humans. If one uses eith- er the rat or mouse metabolic parameters, the extension of the model will predict only how a very large rat or a very large mouse would handle benzene. Therefore, it is essential to have enough information on the human metabolism of a chemical to permit a valid extension of a physiologic model from animals to humans. Comparison of metabolism of xenobiotics by liver 39 slices or cultured cells from laboratory animals and humans should aid in making such extrapolations. CLINICAL TECHNIQUES FOR GATHERING DATA Neither the clinical history of a patient nor information obtained on a population with a standardized questionnaire con- stitutes biologic material in an ordinary sense. However, both can play a role in identifying markers of exposure. Obvious- ly, age is an important biologic deter- minant of disease risk, and the simplest and most efficient way to determine it is to ask. Similarly, symptoms related to altered biologic states that indicate likelihood of disease can be simply asked about. For example, severe respiratory illness before the age of 2 years implies risk of lower respiratory illness at 6- 11 (Samet et al., 1983), and persistent wheezing during childhood predicts dimin- ished pulmonary function in later life (Weiss et al., 1980~. Questionnaire Since the early 1950s, efforts to devel- op standardized procedures for gathering clinical and epidemiologic data on pul- monary health status have been in place (Samet, 1978~. The efforts have been di- rected toward reducing bias and ensuring reliability and validity of the informa- tion obtained. The first recognized standard questionnaire became available in 1960: the British Medical Research Council (BMRC) questionnaire on respira- tory symptoms. The questionnaire was mo- derately revised in 1966 and 1976. The American Thoracic Society (ATS) adapted the BMRC questionnaire in 1968 and pub- lished it with instructions for its use in the United States. Additional modifi- cations have taken place, and the ques- tionnaire has been translated and used extensively throughout the world. In 1978, after an extensive evaluation of the questionnaire, ATS and the Division of Lung Disease (DLD) of the National Heart, Lung and Blood Institute recom- mended a new version (Ferris, 1978~. An

OCR for page 17
40 extensivereviewofthe 1978ATS-DLDques- tionnaire is beyond the scope of this re- port, but we should note that it is effec- tive and reliable for ascertaining respi- ratory symptoms related to the ill effects of cigarette-smoking or other respiratory irritants. Because smoking-habit infor- mation is obtained in a standardized for- mat, the questionnaire makes it possible to measure current and lifetime exposure and allows for comparisons between groups. Specific biologic risk factors iden- tified in the questionnaire for chronic respiratory diseases are summarized in Table 2-4. Other questions can also be used in a clinical setting with patients and provide useful measures of the severi- ty of disease, but are not as well stand- ardized as those discussed. With the use of standardized questions, population groups with different degrees of risk can be defined; the questions can then become useful to delineate biologic markers. Although generally considered effec- tive, the use of questionnaires clearly has limitations. Repeated assessments in what are thought to be stable popula- tions are not without variance' and few (if any) questions have been independent- ly validated (Samet, 1978~. The original survey questions in the ATS-DLD question- naire were designed specifically to identify smoking effects. Few questions are related to specific environmental agents other than cigarette smoke. Now that smoking occurs only in a minority of MARKERS IN PULMONARY TOXICOLOaY subjects, alternative questions might be warranted. And concern has been expressed that better questions need to be developed to deal with symptoms of reac- tive-airway disease, such as asthma. The present questionnaire includes only one question on asthma and one series of questions on wheezing. Developing useful questions on reactive-airway disease will require correlations of responses to new- ly constructed questions with a readily acceptable physiologic test of increased airway reactivity (e.g., nonspecific airway hyperreactivity). Clinical signs and examinations by phy- sicians or other trained observers usually have not proved particularly use- ful as biologic markers of predisease sta- tus. The yellow-stained fingers of a chronic cigarette-smoker might be just as useful "biologic markers" as are ques- tions about smoking habits. Lung Sounds One item in a physical examination that is potentially useful as a biologic marker is the recording of lung sounds (Woolen et al., 1978~. However, instruments for reproducible scaling of lung sounds remain to be developed, and credible scientific investigation of the usefulness of lung sounds has been sparse (Loudon and Murphy, 1984~. Existing technology allows accur- ate recording of sounds transmitted from the airways and parenchyma to the chest wall. Integrating and possibly automating TABLE 24 Biologic Questionnaire Data That Can be Used to Predict Chronic Respiratory Disease in Adults Age Chronic cough and/or phlegm (chronic mucus hypersecretion) Episodes of cough and phlegm Persistent wheeze Dyspnea Smoking history Few cases of emphysema below age 40; risk increases with age; greater risk in men (might be related to greater exposure) Associated with cigarette-smoking; not independently associatedwith risk of COPD; might be associated with excess risk of lung cancer Associated with chronic mucus hypersecretion, increased episodes of pneumonia, and time lost from work Increased risk of asthma; reduced pulmonary function; increased susceptibility to pulmonary irritants Poorly correlated with pulmonary function, but at severe grades associated with reduction and inability to perform pulmonary function test adequately Important predictor of risk (e.g., of lung functional impairment and lung cancer)

OCR for page 17
EXPOSURE the information obtained from those sounds would allow useful categorization of early respiratory injury that could be corre- lated with other markers of exposure. The recording of lung sounds has the advantage of being noninvasive and warrants further exploration. An example of the use of lung sounds is shown in a survey of 270 asbestos factory workers. Shirai et al. (1981) found fine discontinuous lung sounds (i.e., crack- les) more frequently in asbestos workers (32.2%) than in controls (4.5%~. There was good agreement between findings on chest auscultation and sound recordings. In fact, it was found that bilateral basal crackles occurred in asbestos workers before radiographic abnormalities were present. Fine crackles might be valu- able as an early diagnostic marker of pul- monary asbestosis. Respiratory Function The use of respiratory function studies is described in detail in the next chapter. Nonetheless, the summary report by Lippmann ( 1988) related to acute ozone exposures provides an example of the use of clinical techniques to assess exposure in epidemiologic studies. He concluded that there were "progressively increasing functional decrements with each consecu- tive hour of O3 at 0.12 ppm, as well as sub- stantial increase in bronchial reactivi- ty...." Information presented at the U.S.-Dutch meeting "suggests that lung inflammation from inhaled Or has no threshold down to ambient background con- centrations. It was further concluded that rats constitute a good test model for the observed human response to ozone, even though they are less sensitive than humans. Studies in healthy and asthmatic ado- lescents (Koenig et al., 1987) used stand- ard measures of respiratory function- e.g., peak flow, respiratory resistance, and forced expiratory volume (FEY)-and found significant increases in respira- tory resistance in both healthy and asth- matic adolescents after exercise exposure to ozone at 0.18 ppm, but no differences in response between the two groups. Koenig reported that the most important finding 41 is that there was little difference in the effects of ozone or nitrogen dioxide be- tween healthy and asthmatic subjects. Other studies of respiratory responses to ozone exposure in healthy, active chil- dren have also used standard respiratory function measures and found highly sig- nificant changes in PEER in response to changes in ambient ozone concentrations (Spektor et al., 1988~. Imaging Other noninvasive techniques that could be regarded as yielding biologic markers of exposure or disease include radiography and other imaging techniques. Unfortu- nately, the techniques seldom provide evidence of specific exposure or disease. Two notable exceptions are the pneumoconi- oses that constitute specific evidence of dust accumulation in the parenchyma and the pleural reactions that are produced by asbestos exposure; in both cases, the findings must be accompanied by appropri- ate exposure data if they are to be reliable. The major rationale for obtaining screening x-ray pictures has been to iden- tify persons with clinically silent pul- monary tuberculosis. More recently, chest x-ray examination has been used to screen asymptomatic smokers for lung tu- mors and to screen for occult lung or heart diseases in general hospital admis- sions. Clinicians have recommended the use of chest x-ray pictures to establish a baseline for comparison, especially in people presumed to be at risk of lung disease. Even in such select groups, the merits of screening have been debated. Although enthusiasm for chest x-ray pic- tures for screening, in general, seems to be waning, it is worth emphasizing that radiologic examination remains a major diagnostic tool for revealing oc- cupationally induced interstitial lung disease. In the 1970s, the International Labor Organization (ILO) produced a standard- ized procedure for obtaining and reading chest x-ray pictures that allowed crude measurement of exposure to mineral dust (Jacobson and Lainhart, 1972~. Refine

OCR for page 17
42 meets of the procedure in the 1970s have led to a standardized method that increas- es the uniformity and reproducibility of readings of these films made bv cer- tified persons (ILO, 1980~. At least in serious cases, the degree of parenchymal infiltrate is correlated with histopatho- logic evidence of dust accumulation (Sea- ton, 1983~. Thus, the x-ray pictures in specific circumstances become the doc- umentation of the biologic marker of inor- ganic dust exposure. For example, al- though it is less quantitative, the ap- pearance of pleural plaques or pleural thickening on a chest x-ray picture in the presence of a history of asbestos exposure implies a greater likelihood of future asbestos-related disease. Sophisticated imaging techniques used in clinical practice, such as computed axial tomographic (CAT) scanning, can locate and characterize small lesions that could be considered biologic markers of disease. However, the procedures are clinical diagnostic tools and not likely to be used in screening populations of apparently healthy subjects. Such techniques are of tremendous im- portance to experimental medicine and toxicology, because they permit resolu- tion at the subcellular level and might ultimately promote the identification of biologically effective dose, early biologic effect, and altered structure or function. SUMMARY The development of biologic markers of exposure to xenobiotics offers much promise. New molecular biologic tech- niques permit the measurement of such mo- lecular markers as adducts formed with macromolecules in the body; the techniques can be used to detect adducted material in blood, urine, and tissue samples and are sensitive enough for the measurement of adducts formed with DNA or protein in cells washed from the respiratory tract or collected in sputum. Innovative procedures, such as magneto MARKERS IN PULMONARY TOXICOLOGY pneumography, allow estimation of the lung burdens of some types of particles in the lung. Refined histologic techniques have revealed the cellular sites of deposition of inhaled particles in the lung and thus created the potential for calculating the dose to critical cells. Techniques for analyzing markers are well advanced; e.g., new techniques allow analysis of exhaled air, sputum, nasal ravage fluid, and bronchoalveolar ravage fluid for chemical evidence of exposure to specific pollu- tants. Mathematical modeling has advanced to the point where models now include physio- logic measurements, such as blood flow rates, ventilation rates, metabolic rates, and both blood-air and blood-tis- sue partition coefficients. The models have made it much easier to extrapolate from animal pharmacokinetic data to pre- dicted disposition in humans. To make optimal use of the new tech- niques, we must determine the relation- ship between markers of exposure and the characteristics of the exposures that generate them. The markers usually yield only yes-no answers; that is, a particular exposure did or did not occur. But we need to determine the kinetic relationships between formation and breakdown of mark- ers, so that we can use mathematical mod- els to answer the question, "Given this amount of this marker in this tissue, what exposures could have produced the marker?" In addition, there is a need to explore such readily available respiratory tract flu- ids as nasal fluids and sputum for new chemical markers of exposure to specific pollutants. Finally, we must determine the mechan- isms by which environmental pollutants induce lung disease. What are the sites of toxic actions? How much of a given pol- lutant is required at a given site to pro- duce a given toxic response? Knowledge of the mechanisms by which toxicity occurs should provide the most pertinent infor- mation on potential early markers of ex- posure to environmental Pollutants and initial stages of response to them.