Exposure

The first part of this chapter discusses three air sampling methods—total particulate, benzene-soluble particulate fraction, and PAHs, all of which have been used in recent NIOSH investigations to evaluate occupational exposures to asphalt fumes [Almaguer et al. 1996; Hanley and Miller 1996a,b; Kinnes et al. 1996; Miller and Burr 1996a,b, 1998]. Table 4–1 provides a summary of these sampling and analytical methods. In addition, worker exposure data from studies evaluating asphalt refining, hot-mix asphalt plants, road paving, roofing (both manufacturing and installation), flooring, and waterproofing are reviewed and summarized in Appendix B.

4.1 Methods for Analyzing Workplace Air and Dermal Exposures

A variety of sample collection and analytical methods are available for evaluating exposures to asphalt fumes in the workplace. Two methods frequently employed measure either total particulates or the benzene-soluble fraction of total particulates. Unfortunately, neither of these methods measures exposure to distinct chemical components or even a distinct class of chemicals, making it difficult to relate specific components to possible health effects. For example, many organic compounds are soluble in benzene, and any dust or aerosols may contribute to total particulate concentrations. In an attempt to characterize asphalt fumes more accurately, investigators have developed methods to measure individual unsubstituted PAHs, such as acenaphthylene, anthracene, and naphthalene; total PACs; or other potentially irritating substances, such as sulfur-containing compounds. These methods are described below.

4.1.1 Total Particulates as an Indicator of Asphalt Fumes

Total particulates are a measure of all airborne particulates that can be collected on a tared (weighed) sample filter. Several current occupational exposure limits for asphalt fumes are expressed as total particulates. In a study at an asphalt hot-mix plant, the size distribution of approximately 95% to 98% of the asphalt particles was shown to be between 1 and 5 µm in diameter, while at an asphalt paving site, where samples were collected above the screed auger of the paver vehicle, approximately 76% of the particles were between 1 and 5 µm in diameter [Hicks 1995]. These data indicate that asphalt fumes are composed of relatively small particles and may be collected equallywell using the more traditional sampling method (closed-face, 37-mm sampling cassettes) or inhalable samplers. However, further research is warranted to define the various size fractions of asphalt fumes at paving and other worksites where asphalt is used.

In the 1977 criteria document, NIOSH established a recommended exposure limit (REL) of 5 mg/m3 as a 15-min ceiling limit2 for asphalt fumes measured as total particulates. The NIOSH REL was intended to protect workers against acute effects of exposure to asphalt fumes, including irritation of the serous membranes of the conjunctivae and the mucous membranes of the respiratory tract. In 1988, NIOSH (in testimony to the Department of Labor) recommended that asphalt fumes should be considered a potential occupational carcinogen [NIOSH 1988].

Currently, no OSHA standard exists for asphalt fumes. In a 1988 proposed rule on Air Contaminants, OSHA proposed a PEL of 5 mg/m3 as an 8-hr time-weighted average (TWA) for asphalt fume exposures in general industry. This proposal was based on a preliminary finding that asphalt fumes should be considered a potential carcinogen [53 Fed. Reg. 21193 (1988)]. In 1989, OSHA announced that it would delay a final decision on the 1988 proposal because of complex and conflicting issues submitted to the record [54 Fed. Reg. 2679 (1989)]. In 1992, OSHA published another proposed rule for asphalt fumes that included a PEL of 5 mg/m3 (total particulates) for general industry, construction, maritime, and agriculture [57 Fed. Reg. 26182 (1992)]. Although OSHA invited comment on all of the alternatives, its proposed standard for asphalt fumes would establish a PEL of 5 mg/m3 (total particulates) based on avoidance of adverse respiratory effects. The OSHA docket is closed, and OSHA has not scheduled any further action.

The current American Conference of Governmental Industrial Hygienists (ACGIH) threshold limit value (TLV®) for asphalt fumes is 0.5 mg/m3 (8-hr TWA) as a benzene-soluble aerosol (inhalable fraction) or equivalent method with an A4 designation, indicating that it is not classifiable as a human carcinogen [ACGIH 2000]. Irritation is the critical effect.

4.1.2 Benzene-Soluble ParticulateFraction

The benzene-soluble particulate fraction is that portion of total particulates that is soluble in benzene. Organic compounds are generally soluble in benzene, whereas inorganic compounds are not. Historically, this particulate fraction has been used to differentiate between asphalt fumes and other nonasphalt particulates present, such as road dust, at paving sites. Of course, sampling for benzene solubles (or total particulates) assumes that asphalt fumes (as opposed to diesel engine exhaust, for example) are the predominant or sole contributor to air pollution at a worksite. NIOSH Sampling and Analytical Method 5042 contains further details on the collection and analysis of total particulates and benzene solubles.

In the past, because of concerns with the carcinogenicity of benzene, other solvents (such as cyclohexane, acetonitrile, and methylene chloride) have been used in place of benzene to measure the soluble fraction of a particular matrix. When sampling asphalt fumes, however, it is difficult to compare the results because the extraction capability of these solvents varies. For example, carbon disulfide may not be as effective as benzene for extracting the polar compounds in the fumes. NIOSH researchers believe that benzene provides the best overall solubility for asphalt fumes.

4.1.3 Polycyclic Aromatic Hydrocar-bons and Polycyclic Aromatic Compounds

In many asphalt fume studies, researchers have attempted to analyze individual PAHs using either LC/UV/FID or GC/FID. Although this approach has been successful in many matrices containing PAHs, studies of asphalt fumes have shown that these fumes contain a complex mixture of PACs, a class of chemical compounds that contain two or more fused aromatic rings. NIOSH researchers believe that, on an individual basis, these PACs cannot be easily separated or quantified (see section 3.5.3).

In response to this analytical dilemma, NIOSH researchers developed a flow-injection method (NIOSH Method 5800) to measure the total PAC content of asphalt fumes [Miller and Burr 1998, Appendix A]. After it is collected, the asphalt fume sample is extracted from the sampling filter with hexane. This extract is then eluted through a solid-phase extraction column to separate the aliphatic and aromatic compounds. The aromatic compounds are then extracted from the aliphatic compounds using a liquid-liquid extraction procedure.

Because the excitation and emission wavelengths are not the same for all PACs, two sets of excitation and emission wavelengths were used in seven asphalt paving studies conducted by NIOSH [Almaguer et al. 1996; Hanley and Miller 1996a,b; Kinnes et al. 1996; Miller and Burr 1996a,b, 1998]. One set of wavelengths (254-nm excitation, 370-nm emission) is more sensitive for two-ring and three-ring compounds (the lower molecular weight PACs); the second set of wavelengths (254-nm excitation, 400-nm emission) is more sensitive for four-ring and higher compounds (the higher molecular weight PACs). It should be noted that other researchers [Kriech et al. 1999; Kurek et al. 1999] are using similar techniques for measuring the presence or absence of four- to six-ring PACs in asphalt fumes. No occupational exposure limits have been established for total PACs associated with asphalt fumes. NIOSH Sampling and Analytical Method 5800 contains further details on the collection and analysis of PACs.

4.2 Occupational Exposure Data, Air and Dermal Wipe Sam-pling

Comparing historical occupational exposure data from asphalt fume studies can be complicated by many factors, including the complex and variable nature of the asphalt itself, the lack of a single chemical substance accepted as representative of asphalt fume exposure, and the use of different sampling and analyticalmethods. This last factor is important in terms of assessing exposures because such differences can affect what markers are measured for asphalt fume exposure and how comparable the results of different studies are. For example, studies of asphalt fumes may report individual PAHs or total PAHs, but the analytical methods used to obtain results may vary in accuracy and PAH identifications are unreliable (see section 3.5). Also, when solvents other than benzene, such as cyclohexane or acetonitrile, are used to obtain the soluble fraction of total particulates, the results cannot be compared easily because the extraction ability of these solvents varies.

Because of the potential problems encountered when combining results from studies in which sampling and analytical methods differ, environmental data obtained from studies of asphalt refining, hot-mix asphalt plants, road paving, roofing, flooring, and waterproofing are summarized by topic in Appendix B. Analysis of these data indicated that the highest personal total particulate exposures were measured during asphalt flooring and waterproofing activities (1.1 to 42 mg/m3), followed by roofing products manufacturing (0.07 to 15 mg/m3), asphalt refining (0.3 to 14 mg/m3), roofing application (0.04 to 13 mg/m3), activities at hot-mix asphalt plants (0.1 to 7.2 mg/m3), and road paving (0.1 to 5.6 mg/m3). Personal exposures to benzene-soluble particulates followed a similar pattern, with the highest exposures once again being measured during asphalt flooring and waterproofing activities (0.8 to 14 mg/m3), followed by asphalt refining (0.03 to 13 mg/m3), roofing application (0.04 to 6.9 mg/m3), road paving (0.03 to 4.4 mg/m3), and roofing products manufacturing (0.01 to 3.7 mg/m3). The following section discusses several recent asphalt exposure studies in greater detail.

4.2.1 NIOSH/FHWA Evaluation of Asphalt Paving Workers

Between 1994 and 1997, seven surveys [Almaguer et al. 1996; Hanley and Miller 1996a,b; Kinnes et al. 1996; Miller and Burr 1996a,b, 1998] were completed as part of an interagency agreement between NIOSH and the Federal Highway Administration (FHWA) of the U.S. Department of Transportation. The objectives were to (1) develop and field test new methods of characterizing asphalt fume exposures and (2) identify potential health effects associated with asphalt exposures (health effects are discussed in section 5).

At each NIOSH survey site, full-shift personalbreathing-zone samples were collected from the paving crew, which typically consisted of six to 10 workers, for total particulates and the benzene-soluble particulate fraction. Table 4–2 shows that average personal-breathing-zone air concentrations for both total particulates and benzene solubles were below 0.5 mg/m3, TWA. Table 4–3 shows personal-breathing-zone air concentrations for PACs that were collected and analyzed using a method similar to NIOSH Sampling and Analytical Method No. 5506 (see Table 4–1 and section 3.5.3 for more information). Two spectrofluorometric emission wavelengths were used in the PAC analyses. These were 370 nm, which is more sensitive to the lower molecular weight, two–ring and three–ring PAC compounds (termed PAC370); and 400 nm, which is more sensitive for the higher molecular weight, four–ring and larger compounds (referred to as PAC400). In these studies, concentrations of PAC370 always exceeded concentrations of PAC400, implying that the lower molecular weight, two- and three-ring PACs (postulated by NIOSH investigators to be more responsible for irritant effects) may be more abundant in asphalt fumes.

In addition to the personal-breathing-zone samples, area air samples were collected over the screed auger section of the paver vehicle and analyzed for total particulates, respirable particulates, benzene solubles, and total hydrocarbons. Area air samples were also collected for VOCs, carbon monoxide, hydrogen sulfide, sulfur dioxide, and ozone, substances which NIOSH investigators theorized could also be present during road paving.

Area air sampling results for respirable and total particulates, benzene-soluble particulates, and total hydrocarbons are summarized in Table 4–4. Across the seven paving sites, area concentrations of respirable particulates at the screed auger ranged from 0.055 to 0.97 mg/m3 , total VOCs (measured as either n-hexane or Stoddard solvent) ranged from 0.5 to 30 mg/m3 , and concentrations of selected individual VOCs (benzene, toluene, xylene, and methyl isobutyl ketone) were generally less than 1 part per million (ppm). At some sites, area concentrations of carbon monoxide ranged up to 1,000 ppm where gasoline–powered equipment, such as vibrating tampers or portable generators, was in use. At all survey locations, concentrations of hydrogen sulfide and sulfur dioxide were not detected.

4.2.2 NIOSH Evaluation of Asphalt Paving among Tunnel Workers

In 1996, NIOSH evaluated exposures of paving crews working within the Third Harbor Tunnel in Boston, MA [Sylvain and Miller 1996]. The work included the collection of full-shift personal-breathing-zone and area air samples for total particulates and benzene solubles, questionnaires administered to obtain information on symptoms, and tests of peak lung flow (see section 5.1 for details on the medical results). As shown in Table 4–5, personal exposures to total particulates and benzene solubles averaged 1.6 and 0.76 mg/m3, respectively. These concentrations were up to three times higher than exposures measured during the seven NIOSH/FHWA surveys at open-air roadway paving sites (see section 4.2.1). Poorer ventilation in the tunnel (as compared to open-air paving sites) likely contributed to these higher personalbreathing-zone exposures.

4.2.3 Cross-Sectional Occupational Exposure Assessment Study

In a cross-sectional occupational exposure assessment [Hicks 1995] covering road paving sites, hot-mix plants, refineries and terminals, roofing manufacturing plants, and roofing application sites, 219 full-shift personalbreathing-zone air samples were collected and analyzed (most sampling periods ranged from 7 to 9 hours) [Hicks 1995]. In addition to air samples, 131 dermal wipe samples for benzene solubles were collected (see section 4.3). The objective of this study was to characterize worker exposures to asphalt fumes via both airborne and dermal routes.

As shown in Table 4–6, concentrations of total particulates (0.18 to 1.4 mg/m3) and benzene solubles (0.15 to 0.27 mg/m3) varied across all industry types. Geometric mean exposures in all sectors were comparable when measured as ben-zene solubles. The highest concentration for to-tal particulates was at roofing manufacturing plants, but this result may be attributable to nonasphalt-related particles in these plants. The air samples with the most abundant PAH com-pounds were obtained from workers at construction sites (roofers and paving crews) (Table 4–7). Hicks reported that lower molecular weight PAHs, such as naphthalene, were more frequently detected than the higher molecular weight compounds, such as B(a)P. Carcino-genic PAHs (chrysene, B(a)P, and benzo(b) fluoranthene) were detected in personal samples collected from employees working in the in-dustry categories of refineries and terminals, roofing manufacturing, roofing contractors, and paving operations. Fluorene, naphthalene, and phenanthrene were detected in all of the industry categories. It should be noted that HPLC with an ultraviolet/fluorescence detector (the method used in the Hicks study) may not be able to distinguish discrete PAHs present in asphalt fumes. See section 3.5 for a more complete discussion of the analysis of asphalt fumes.

4.2.4 Exxon Cross-Sectional Evaluation of Asphalt Workers

Personal exposures and health outcomes of 170 workers in five segments of the asphalt industry (hot-mix plants, terminals, roofing manufacturing plants, roofing application sites, and paving sites) were evaluated by Exxon Bio-medical Sciences [Exxon 1997]. Gamble et al. [1999] have published a summary of these data. Personal samples were collected across two work days for total and respirable particulates and the benzene-soluble fraction of total par-ticulates. Samples were also collected for VOCs, nitrous oxide, hydrogen sulfide, sulfur dioxide, and ozone. The health outcomes measured in-cluded changes in lung function between shifts and the administration of a questionnaire on symptoms (see section 5.1.3).

Full-shift personal-breathing-zone concentra-tions ranged up to 6.2 mg/m3; respirable par-ticulates up to 1.4 mg/m3; benzene solubles up to 1.3 mg/m3; and VOCs up to 20 mg/m3 . Table 4–8 summarizes the geometric means and maximum exposures to total particulates, re-spirable particulates, benzene-soluble par-ticulates, and VOCs by industry. Concentrations of nitrous oxide, hydrogen sulfide, and sulfur di-oxide were typically near or below detection limits of the analytical methods used. Ozone concentrations were below 100 parts per billion.

4.2.5 Occupational Dermal Exposures

Dermal exposure to asphalt fumes has been ex-amined using skin wipes, which represent the potential contribution of dermal exposure to total body burden. Wolff et al. [1989] collected 10 skin-wipe and nine personal-breathing-zone samples from 10 roofers who had removed an old coal-tar-pitch roof and replaced it with an asphalt roof. PAHs were detected in samples from the breathing zones of employees involved with applying asphalt on two separate days (5.8 and 22 µg/m3, mean) and removing coal-tar pitch (9.6 and 23 µg/m3 , mean). Because NIOSH Sampling and Analytical Method No. 5506 was used, the PAH identifica-tions and the concentration data are considered to be unreliable; however, these data are included for completeness. PAH residues per square centimeter of skin were higher in postshift samples (6.1 to 31 nanograms per square centimeter [ng/cm2]) than in preshift samples (0.44 to 2.2 ng/cm2). Eight of nine cases showed a significant correlation (r=0.97) be-tween PAHs found in personal air samples and in postshift skin wipe residues. However, employees monitored during the entire roofing application were potentially exposed to PAHs during both the removal of the old coal-tar-pitch roof and the application of hot asphalt for the new roof.

One-hundred thirty-one postshift dermal wipe samples were collected from workers at refin-eries, hot-mix facilities, paving sites, roofing manufacturing plants, and roofing sites and analyzed for PAHs in an exposure assessment study sponsored by the Asphalt Institute [Hicks 1995]. These samples were obtained by wiping the foreheads or backs of hands of selected workers with premoistened smear tabs and then analyzing the wipes for the PAH species listed in Table 4–9.

The PAH concentrations determined from these postshift samples ranged from 2.2 to 520 ng/cm2 . Employees in paving operations produced the largest number of PAHs detected (12), while refinery and roofing installation workers had the fewest (2). Naphthalene was detected at all sites. Table 4–10 shows the six PAHs (of the17 PAHs analyzed within this sample set) that were above the detection limit.

NIOSH investigators have collected preshiftand postshift skin wipe samples during paving operations at three separate locations[Zey 1992a,b,c]. The samples were analyzed as described in Wolff et al. [1989]. No PAHs were detected in any of the skin wipe samples, which may have been because of low concentrations of asphalt fumes during paving operations and because the PAH concentrations were below the detection limit of the analytical method.

In Zhou [1997], pre-and postshift hand wipes were collected from a group of 17 asphalt-exposed road pavers and 16 controls to evaluate dermal PAH exposure. These hand-wipe samples were analyzed by HPLC fluorescence for the following nine PAHs: anthracene, fluoranthene, pyrene, benzo(b)fluoranthene, benzo (k)fluoranthene, dibenz(a,h)anthracene, benz(a) anthracene, chrysene, and B(a)P. Zhou reported that among the group exposed to asphalt, total PAH, carcinogenic PAH, and pyrene concentrations increased when pre- and postshift hand-wipe samples were compared (Table 4–11.)

*All samples were analyzed for these PAH species; 20% of these sampleswere also analyzed for the remaining PAHs.
NOTE: These skin wipe samples were analyzed using HPLC fluorescence. We have included these data for completeness; however, because of theasphalt fume matrix and the analytical technique used to evaluate the PAHs, the PAH identifications and the concentration data are considered to be unreliable. See section 3.5.3 for more information.

4.2.6 Summary

Based on results of studies of open-air paving sites, refineries, asphalt distribution terminals, and hot-mix asphalt plants, mean personal airborne exposures to asphalt fumes were generally below 1.0 mg/m3 for total particulates and 0.3 mg/m3 for benzene solubles, calculated as a full-shift TWA (Table 4–12). Full-shift TWA personal exposures measured during some activities, such as underground paving, roofing manufacturing, and roofing application, however, were higher, ranging up to 1.6 mg/m3 for total particulates and up to 0.76 mg/m3 for benzene solubles.

While PAH data were included for completeness, the results are not provided in this summary because PAH identifications and concentration data are considered to be unreliable. See section 3.5.3 for more information.

4.3 Biomarkers

In addition to measures of ambient exposures to occupational chemicals, various studies have used readily accessible body fluids and/or physiological functions as biomarkers for exposure to asphalt fumes. Urinary thioether excretion, glucaric acid metabolites in urine, detection of mutagens in urine, sister chromatid exchange and primary DNA damage in lymphocytes, urinary 1-hydroxypyrene, and DNA or protein adducts have been described as indicators of exposure to or effects of asphalt fumes.

4.3.1 Urinary Thioethers

Urinary thioethers have been proposed as potential biomarkers of internal exposure to electrophilic compounds [Van Doorn et al. 1981]. The glutathione-S-transferase (GST) enzyme system facilitates the conjugation of glutathionewith electrophilic agents. This conjugation step usually results in detoxification of the agent and enhances its elimination in bile or urine. In addition to xenobiotic agents, numerous endobiotic ma-terials are also conjugated to glutathione by GST. When compounds are conjugated with glutathione, mercapturic acids and other thioethers appear in the urine as nonspecific indicators of exposure to electrophilic agents.

Numerous researchers have attempted to correlate asphalt exposure (both road paving and roofing operations) to increased urinary thioether excretion [Lafuente and Mallol 1987; Burgaz et al. 1988, 1992; Pasquini et al. 1989; Hatjian et al.1995a, 1997]. These efforts have been unsuccessful. Even in limited cases where potential correlations may have existed, values were within normal human ranges.

4.3.2 Urinary Glucaric Acid

Glucaric acid excretion is another indirect measure of exposure to materials eliminated by conjugation. Like thioethers, many endo-biotic agents are conjugated with glucuronic acid for transport and elimination by orga-nisms. In theory, increased exposure to agents that are made less toxic via glucuronidation should result in increased elimination that could be estimated by glucaric acid excretion.

Pasquini et al. [1989] and Hatjian et al. [1995a, 1997] measured D-glucaric acid in workers exposed to asphalt fumes. They reported no differences in concentrations of urinary D-glucaric acid in exposed workers compared to unexposed workers.

4.3.3 Mutagenic Activity in Urine

Mutagens excreted in the urine are thought to be indicative of exposure to mutagenic agents. The presence of mutagens excreted in urine of asphalt-exposed workers involved in road paving was examined by Pasquini et al. [1989]. Results were based on the Ames Salmonella mutagenicity assay with TA98 strain and metabolic activation by rat S9. In nonsmoking individuals, asphalt-exposed workers had a significant increase in mutagenic activity in urine when compared to unexposed workers. However, among smokers, there was no significant difference in mutagenic activity in urine between exposed and unexposed workers, hence Pasquini et al. could not attribute this activity to asphalt exposure with confidence.

4.3.4 Sister Chromatid Exchange

Sister chromatid exchange (SCE) is a sen-sitive, indirect measure of genetic damage. However, SCE provides no information as to the identity of the genotoxic agent. SCE in white blood cells has been used as a bio-marker to estimate genotoxicity of asphalt exposure. Hatjian et al. [1995b] reported SCE frequencies for a combined group of road pavers and roofers and concluded that the mean SCE frequency was increased (P<0.05) in the paver-roofer group compared to un-exposed office workers. However, these mean SCE frequency levels did not differ from those of a group of manual workers with no asphalt exposure. The office workers were allnonsmokers, while other groups included 20% or fewer smokers.

In a second report, Hatjian et al. [1995a, 1997] combined individuals from the first study [Hatjian et al. 1995b] with new groups of workers. They were divided into four occupa-tional groups: office workers, manual laborers, two groups of road pavers, and two groups of roofers. One group of pavers and both groups of roofers had significantly higher SCE frequencies than either manual laborers or office workers.

4.3.5 Urinary 1-Hydroxypyrene

Urinary 1-hydroxypyrene is often used as a biomarker of exposure to pyrene and, by ex-trapolation, to PAHs from any source [Lauwerys and Hoet 1993]. Table 4–13 summarizes the reported use of urinary 1-hydroxypyrene as a biomarker of exposure to asphalt and asphalt fumes. The logic behind this use is that asphalt and asphalt fumes are complex mixtures containing PAHs, including pyrene. Pyrene in exposed humans is metabolized to 1-hydroxypyrene and excreted in urine, mainly as glucuronide. For biological monitoring, postshift urine specimens are col-lected and analyzed; the analytical methods call for hydrolysis of glucuronide and other conjugates of 1-hydroxypyrene before the liberated 1-hydroxypyrene is quantified. The concentration of 1-hydroxypyrene frequently is normalized to the concentration of creat-inine to correct for urine dilution.

The data in Table 4–13 are grouped by oc-cupation and, within the road paver group, by decreasing mean level of urinary 1-hydroxy-pyrene. The ranges of concentrations found for the occupationally exposed and reference popu-lations overlapped, and in many cases, the differences between mean concentrations for the two populations were not statistically significant. This result most likely reflects the contributions by nonoccupational exposures to PAH, such as ambient air pollution, tobacco smoke, and fried, roasted, and charbroiled food. The influence of nonoccupational exposures is demonstrated by mean urinary 1-hydroxypyrene concentrations in the populations of road pavers, which varied over 60-fold and overlapped the 50-fold range of mean concentrations of the reference populations. The data in Table 4–13 are more easily compared after nonoccupational exposures are adjusted by dividing the means of the exposed populations by the means of the reference populations. For road pavers, these ratios were 2.2 to 3.7; for roofers, 1.3 and 2.7; and for workers with asphalt, 1.4 and 1.8. In comparison, road pavers working with surfacingmaterial containing a mixture of asphalt and coal tar had mean concentrations 3.3 to 12 times the mean for the reference population [Jongeneelen et al. 1988].

These observations suggest that the utility of urinary 1-hydroxypyrene as a biomarker of exposure to asphalt and asphalt fumes is limited. This biomarker may prove useful for revealing relatively high exposures to asphalt and for demonstrating reduction of exposures resulting from implementation of engineering controls and improved work practices. However, there are several weaknesses in the use of 1-hydroxypyrene as an indicator of PAH exposure. For instance, nonoccupational exposures to PAH may complicate determination of the contributions of low-level occupational exposures. Moreover, in the case of exposure to fumes, the fraction of pyrene in the fumes will vary with the concentration of pyrene in the asphalt and the temperature of the bulk asphalt, both factors that decrease how accurately urinary 1-hydroxypyrene represents overall exposure.

These observations suggest that the utility of urinary 1-hydroxypyrene as a biomarker of exposure to asphalt and asphalt fumes is limited. This biomarker may prove useful for revealing relatively high exposures to asphalt and for demonstrating reduction of exposures resulting from implementation of engineering controls and improved work practices. However, there are several weaknesses in the use of 1-hydroxypyrene as an indicator of PAH exposure. For instance, nonoccupational exposures to PAH may complicate determination of the contributions of low-level occupational exposures. Moreover, in the case of exposure to fumes, the fraction of pyrene in the fumes will vary with the concentration of pyrene in the asphalt and the temperature of the bulk asphalt, both factors that decrease how accurately urinary 1-hydroxypyrene represents overall exposure.

4.3.6 DNA Adducts

One of the primary hypotheses of chemicacarcinogenesis is that the interaction betweespecific chemicals (or their metabolites) anDNA can result in damage to DNA that malead to neoplastic cells or cancer [Randerath eal. 1983; Slaga 1984]. The chemically modified DNA is referred to as adducted DNA omore simply, DNA adducts. Reactive chemicals may also bind to cellular proteins, thuforming protein adducts. DNA and proteiadducts in readily accessible tissues have beeused as biomarkers of biological effects iworkers exposed to asphalt.

Herbert et al. [1990] used the 32P postlabeling methodology to examine DNA adducts in the white blood cells of 12 roofers with asphalt exposure and 12 unexposed individuals matched for age, sex, and smoking status. They also performed tests for PAH on personal-breathing-zone samples and skin wipes using HPLC fluorescence (NIOSH Method 5506). Eighty-three percent of the roofers, compared to 17% of the unexposed individuals, had detectable concentrations of aromatic DNA adducts. DNA-adduct con-centrations were not correlated with PAH content of personal-breathing-zone samples, but were positively associated with postshift skin concentrations of PAHs. The contribution of exposure to PAH from removing an old pitch roof (type of pitch not specified) could not be separated from exposures while ap-plying a new asphalt roof. It must also be noted that in the two roofers evaluated for adduct type, the adducts did not appear to be the major one normally associated with B(a)P exposure. These data indicate that B(a)P is not the major source of DNA-adduct formation, but that a yet-to-be-identified compound(s) contributes to DNA-adduct formation.

Lee et al. [1991] used an immunoassay to measure protein adducts in serum albumen in the same group studied by Herbert et al. [1990] (12 roofers and 12 unexposed indi-viduals). These researchers reported sig-nificantly greater borderline numbers of ad-ducts in exposed workers compared to unexposed individuals (P #0.10). The anti-body used in these studies reacts with adducts of B(a)P and, to different degrees, cross-reacts with several other PAHs. However, because the identity of the adducts being measured was unknown, this immunologic assay may not provide an accurate estimate of adduct con-centrations.

Fuchs et al. [1996] measured primary DNA damage (strand breaks) and DNA adducts in mononuclear cells of workers exposed to asphalt.³ These workers included roofers (n=7), pavers (n=18), and asphalt painters (n=9). The control group (n=34) consisted of students and office workers. All roofers and 10 members of the control group smoked. The roofers studied had significantly greater (P<0.002) numbers of DNA strand breaks, and these were found to increase during the work week. Because the type of roofing work and materials used were not defined, exposure to coal tar could not be excluded. Pavers and asphalt painters did not differ statistically from controls in the incidence of DNA strand breaks; however, the numbers of strand breaks were found to increase during the work week in the group of pavers. DNA adducts were found in 10 of 14 samples obtained from pavers and asphalt painters, and DNA adduct concentrations were positively correlated with age and years of exposure. Technical problems prohibitedanalysis of DNA adducts in other subjects.

Zhou [1997] measured DNA adducts in exfoliated uroepithelial cells in 12 road pavers and 13 road construction workers who had no exposure to asphalt fumes. No correlation was found between exposure to asphalt fumes and DNA adducts.

4.3.7 Conclusions

Biomarker studies conducted in workers exposed to asphalt are summarized in Table 4–14. Until a chemical component specific to asphalt fumes is identified, a biomarker specific and unique to asphalt exposure cannot be developed. Many of the studies of biomarkers conducted in workers exposed to asphalt were designed to determine if exposure to PAHs had occurred [Burgaz et al. 1992; Hatjian et al. 1995a,b, 1997; Herbert et al. 1990; Lee et al. 1991; Levin et al. 1995; Boogaard and van Sittert 1994, 1995; Zhou 1997]. Other studies utilized endpoints that were not specific to PAH exposure [Lafuente and Mallol 1987; Burgaz et al. 1988, 1992; Pasquini et al. 1989; Hatjian et al. 1995a,b, 1997].

Exposure to potentially genotoxic compounds may occur during work with asphalt. Evidence of such exposures is indicated by—

• The observed concentrations of mutagens in the urine of asphalt workers who were nonsmokers [Pasquini et al. 1989]
• Observed concentrations of SCE [Hatjian et al. 1995a,b, 1997] and DNA adducts [Herbert et al. 1990] in white blood cells
• Exposure to PAHs as detected by the presence of the sentinel urinary metabolite, 1-hydroxypyrene [Burgaz et al. 1992; Levin et al. 1995; Zhou 1997].

However, in the studies reported, smoking, environmental factors, and diet frequently confounded study interpretations. It is difficult to categorize exposure based solely on occupational classification, because exposures may have been misclassified. In every case, the significance of the relatively small differences in biomarkers observed in exposed workers compared to controls was not clear.