6
RISK CHARACTERIZATION OF EXPOSURES TO ZINC CADMIUM SULFIDE

RISK IS THE PROBABILITY of a specific outcome (such as an adverse health effect) under a particular set of conditions. Human-health risk assessment entails the evaluation of toxicological and related information on the environmental agents and the extent of human exposure to those agents. The product of the evaluation is a statement regarding the probability that populations so exposed will be harmed and to what degree.

A 1994 NRC report describes a 4-step analytic process for performing a human-health risk assessment, which involves hazard identification, dose-response assessment, exposure assessment, and risk characterization. When a substance leaves a source (such as an industrial facility), it moves through an environmental medium (such as the air), and results in an exposure (for instance, people breathe air that contains the substance). The exposure creates a dose in the exposed people (the amount of the chemical entering the body, which may be expressed in any of several ways). The magnitude, duration, and timing of the dose determine the extent to



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 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests 6 RISK CHARACTERIZATION OF EXPOSURES TO ZINC CADMIUM SULFIDE RISK IS THE PROBABILITY of a specific outcome (such as an adverse health effect) under a particular set of conditions. Human-health risk assessment entails the evaluation of toxicological and related information on the environmental agents and the extent of human exposure to those agents. The product of the evaluation is a statement regarding the probability that populations so exposed will be harmed and to what degree. A 1994 NRC report describes a 4-step analytic process for performing a human-health risk assessment, which involves hazard identification, dose-response assessment, exposure assessment, and risk characterization. When a substance leaves a source (such as an industrial facility), it moves through an environmental medium (such as the air), and results in an exposure (for instance, people breathe air that contains the substance). The exposure creates a dose in the exposed people (the amount of the chemical entering the body, which may be expressed in any of several ways). The magnitude, duration, and timing of the dose determine the extent to

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests which the toxic properties of the substance are realized in exposed people (the risk). The subcommittee reviewed toxicological and related information on ZnCdS and cadmium compounds and a vast amount of exposure data from the Army's ZnCdS dispersion tests conducted in various US and Canadian locations. Results of the subcommittee's analyses can be found in Chapters 2, ''Input from the Public''; 3, "Toxicity and Related Data on Zinc Cadmium Sulfide"; 4, "Toxicity and Related Data on Selected Cadmium Compounds"; and 5, "Exposure Assessment." Greater detail on exposure data can be found in Appendix B, "A Summary of Doses and Concentrations of ZnCdS Particles from Army's Dispersions Tests." This chapter combines information from the hazard-identification and dose-response assessments, as presented in Chapters 3 and 4, with the exposure-assessment, as presented in Chapter 5, to determine what magnitude of human exposure to ZnCdS might produce adverse health effects. The noncancer and cancer risk assessments of exposures to ZnCdS are discussed below. RISK ASSESSMENT FOR NONCANCER HEALTH EFFECTS Establishing health risk for exposures to agents that might be associated with noncancer health effects is based on the concept that there is a level of exposure below which no adverse health effect would be expected to occur—the so-called threshold dose. The quality of such a prediction is highest when it is based on epidemiologic or clinical studies carried out under conditions that most closely mimic the exposure situation being evaluated. When human-exposure information is not available, one must attempt to extrapolate from the best existing database, often using animal studies. One must recognize possible differences in dose, exposure route, duration, species, and chemical and physical properties of the test material. Risk assessors usually assume that humans and the most sensitive animal species are equally sensitive to the test chemical unless evidence indicates otherwise. The most scientifically defensible assessments are

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests based on the best available information from a combination of human and experimental-animal studies. Accuracy and precision in estimating noncancer human health risk are largely determined by the quality and quantity of the toxicologic data available for analysis. A wide variety of noncarcinogenic responses can be potentially produced by exposure to chemical substances. The responses can be minor and temporary or severe and permanent. The nature and severity of a health effect can depend on several factors, such as the amount of the chemical to which a person is exposed and the potency or toxicity of the chemical itself. Other factors peculiar to the exposed person can influence the effect, including age, sex, health status, lifestyle, and diet. The health effects depend on the route by which the chemical enters the body, the changes in the chemical as it moves through the body, and the specific target organ that is the most susceptible. Noncancer risk assessment normally incorporates a number of "safety" or "uncertainty" factors, which are applied when there is a need to accommodate human response variability, including response in sensitive subgroups; to predict human response from animal data; to extrapolate from subchronic to chronic exposure; to predict a no-observed-adverse-effect level (NOAEL) from a lowest-observed-adverse-effect level (LOAEL); or to use a database that is considered incomplete. Factors ranging between 1 and 10 are most often used for each of the sources of uncertainties listed above, depending on the nature and sensitivity of the adverse effect in question. As the data become more uncertain or unreliable, higher safety factors are applied. Any toxic response might be used for establishing the NOAEL or LOAEL as long as it is the most-sensitive toxic effect and is considered likely to occur in humans. The exposure limit for noncarcinogenic health effects is usually derived from the animal-test data that leads to a NOAEL. This standard approach seems appropriate and was used by the subcommittee for assessing the potential noncancer health risk associated with exposure to ZnCdS. The toxicity database on ZnCdS is not adequate to be used in risk estimates for human exposure. Therefore, toxicity information on some other related chemical can be used to estimate the risk associated with ZnCdS exposures. The subcommittee considered three possible scenarios to esti-

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests mate the noncancer risk associated with exposure to ZnCdS from the Army's dispersion tests. In the worst-case scenario, ZnCdS would have the toxic properties of soluble cadmium compounds. However, the physical and chemical properties of ZnCdS are known. It is insoluble in water and lipids and only poorly soluble in strong acids. If it is assumed that in vivo ZnCdS becomes as soluble as any cadmium compound and might release its cadmium ions to react freely with biologic targets, it would be appropriate to estimate the toxicity of ZnCdS from the toxicity of soluble cadmium compounds in general. It is extremely doubtful whether such an assumption is warranted. Our general understanding of metal toxicology gives this worst-case scenario little, if any, plausibility. In a second scenario, ZnCdS might actually be a biologically inert particle, not more toxic than all the other respirable particles present in our daily environment. This scenario is based on physicochemical properties of ZnCdS (only poorly soluble in strong acids), which results in a lack of acute oral or dermal toxicity. However, the lack of a comprehensive evaluation of the toxicity of ZnCdS, particularly long-term low-level effects, and the fact that ZnCdS might be degraded to some extent, albeit only very slowly, preclude endorsement of such an assumption. In the third scenario, ZnCdS would have toxic properties similar to those of cadmium sulfide, CdS, an insoluble cadmium compound. ZnCdS has a crystalline structure similar to that of CdS; the only difference is that in ZnCdS, zinc replaces 80% of the cadmium in the lattice. CdS is insoluble in water and lipids and slightly soluble in strong acids. Experimental data on toxicokinetics and toxicity of CdS are available. They clearly show that the cadmium in CdS is much less bioavailable than the cadmium in soluble compounds. Furthermore, it is plausible that the fusion of CdS with zinc sulfide, ZnS, at 900ºC will produce a crystalline lattice structure from which cadmium is even less bioavailable than CdS. Until experimental data to the contrary become available, it may reasonably be assumed that the fusion of CdS with ZnS essentially preserves the physicochemical and hence toxic properties of CdS. Therefore, the subcommittee chose to base its assessment of the potential toxicity of ZnCdS for noncancer health effects on the toxicity of CdS. Results of studies conducted by Oberdörster (1990) and Oberdörster

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests and Cox (1990) indicate that CdS inhaled into the lung is not solubilized and absorbed into the blood for distribution to other organs. Therefore, the subcommittee concluded that evidence suggests that sporadic exposures of the public to small amounts of ZnCdS should not result in any systemic toxicity (toxicity to the kidney, bone, immune system, or reproductive and developmental system). The only organ system considered to be potentially at risk because of exposure to particles of ZnCdS was the respiratory tract; the particles of ZnCdS were respirable and would be expected to deposit deep in the lungs. Results of studies of Oberdörster and others (1985) and of Glaser and others (1986) indicate that CdS has little toxicity in the respiratory tract. The subcommittee used the studies of Glaser and others (1986) as the basis for their risk assessment because the cumulative exposures were higher in that study. In the Glaser and others (1986) studies, rats exposed to CdS at 39,600 mg-min/m 3 had only a mild pulmonary response. Using those data and dividing by an uncertainty factor of 10 to extrapolate from a LOAEL to a NOAEL, by a factor of 10 to extrapolate from animal to human exposures, and by another factor of 10 to account for sensitive populations, one would not expect adverse health effects, even in sensitive populations, from exposure to cadmium sulfide in an insoluble form, such as ZnCdS, at 39.6 mg-min/m3 or 39,600 µg-min/m3. This level of 39,600 µg-min/m3 CdS (inhaled dose of 594 µg) or 30,900 µg-min/m3 cadmium (39,600 x 0.78 = 30,900) or 513 µg of inhaled cadmium dose (30,900 µg-min/m3 x 0.0166 m3/min = 513 µg) is considered to be a "safe" level of exposure and was compared with the maximal inhaled doses at the test sites. The four highest estimated inhaled cadmium doses were 6.8 µg (Minneapolis), 24.4 µg (St. Louis), 14.5 µg (Winnipeg), and 390 µg (Biltmore Beach, FL). Inhaled cadmium doses were derived by multiplying the maximal exposures to ZnCdS (expressed as µg-min/m3 and shown in Table B-1 of Appendix B) by 0.0166 3m /min (the volume of air inhaled by an active person in 1 min). The product is then multiplied by 0.156 (the proportion of cadmium in ZnCdS). For example, the maximal ZnCdS exposure in Biltmore Beach, FL was 150,000 µg-min/m3. Multiplying this by 0.01663m /min yields an inhaled dose of ZnCdS of about 2,500 µg. Multiplying that by 0.156 yields an inhaled cadmium dose of about 390 µg. Thus, inhaled cadmium doses from

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests ZnCdS tests did not exceed the "safe" level (513 µg) at any test site. On the basis of the above information, the subcommittee concludes that the exposures to ZnCdS from the Army's tests (for which data are available) did not increase people's exposure to cadmium over background levels by enough to have caused noncancer health effects in exposed persons. RISK ASSESSMENT FOR CANCER The cancer risk associated with a particular chemical is estimated by multiplying the total exposure to it by its cancer-potency factor. 1 The subcommittee based its cancer risk estimates on the cadmium content of ZnCdS because no studies on the carcinogenicity of ZnCdS have been conducted and cadmium is the most-toxic component of ZnCdS. As stated previously, this approach will therefore overestimate the risk. Assuming that 100% of the cadmium is bioavailable, estimates of lung-cancer risk were calculated for all test sites for which air concentrations are available based on the maximal recorded cumulative exposures of ZnCdS in the air and the carcinogenic potency derived from a study of workers exposed to cadmium by inhalation and oral ingestion of dust containing cadmium compounds. Table 6-1 summarizes the assumptions made in calculating the cancer risk estimates. The cancer risk estimates are based on the total cumulative exposure to cadmium in ZnCdS. As will be shown later, the cancer-potency risk factor was derived from the occupational exposure of workers to cadmium. That provided an estimate of the human potency factor expressed as lung-cancer risk per milligram of cadmium. Because workers were exposed to cadmium compounds at higher air concentrations for periods much longer than those of any people exposed from the ZnCdS releases, this cancer-potency risk factor might overestimate the risk for much shorter exposures to ZnCdS. 1   Cancer-potency factor is a value used by regulatory agencies to describe the inherent potency of carcinogens. The factor is derived with dose-response models. Upper limits on cancer-potency factors, also called q1*, are used to estimate an upper limit on the likelihood that lifetime exposure to a particular chemical could lead to excess cancer deaths.

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests TABLE 6-1 ZnCdS Cancer Risk Estimates: Major Assumptions Made and Their Effects on Lung-Cancer Risk Estimates Assumption or Condition Impact 100% of cadmium in ZnCdS was bioavailable If only a fraction of cadmium is bioavailable, all risks are overestimated Cancer potency of cadmium based on exposures of adult workers; cancer-risk estimates multiplied by 10 to account for possible higher sensitivity in some subpopulations, such as the young or elderly Overestimates cancer risks by a factor of 10 for most exposed people Short-term exposure results in a cancer risk proportional to the duration of exposure (total dose) Could overestimate cancer risk by an unknown margin Sampling time was adequate to cover the total airborne exposure for a test release If sampling time were inadequate, dose could have been underestimated The best estimate of the cumulative exposure at the maximal-exposure location is used; the best estimate is likely to be within a factor of 2 of the maximal cumulative exposure at any fixed location Increases the estimate of risk because of the premise that an individual remains at the maximal-exposure location for an extended period Conservative assumptions regarding lung-cancer potency and dose of ZnCdS were used, with maximal recorded cumulative exposures at a test site, so upper bounds of lung-cancer risk were estimated It is likely that the lung-cancer risk is overestimated for most, if not all, people; the cancer risk might approach zero Traditionally, the cancer-potency factor and exposure are both expressed on the basis of average daily exposure over a lifetime. That does not imply here that a lifetime exposure is required to produce cancer. As used here, it is only a dosimetric convention that does not affect the estimation of risk. The average daily exposure is directly proportional to the cumulative exposure on which the cancer risk estimate is based. The cancer risk estimates were calculated first on the basis of a lifetime of exposure and then reduced to the proportion of a lifetime during which people were actually exposed in the test areas. This is equivalent to using the daily exposure averaged over a lifetime. The procedure of multiplying

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests the risk for a lifetime exposure by the fraction of actual exposure is based on calculations with the multistage model of carcinogenesis (Kodell and others 1987) and the Moolgavkar-Knudson-Venzon 2-stage clonal-expansion model (Chen and others 1988), which show that it is not likely to underestimate the risk associated with short-term exposure by more than a factor of 10. That was verified by a comparison of tumor risks related to lifetime and less-than-lifetime exposures reported for animal studies (Gaylor 1988). Cancer risk estimates were multiplied by 10 to account for exposure during possibly sensitive ages, such as childhood. It is unlikely that the procedures used underestimated the lung-cancer risks associated with the reported exposure to cadmium (in ZnCdS) from the Army tests. LUNG-CANCER POTENCY OF CADMIUM Thun and others (1991) report excess lung-cancer risk in workers exposed to various amounts of cadmium by inhalation, namely, differential risks from a standard population of -1.77%, +2.87%, and +14.23% for median exposures of 767; 3,315; and 11,507 µg-yr/m3, respectively. They offer several possible reasons for the reduction in lung cancer for the low-exposure group, including the possibility of a select healthy-worker effect. Regardless, the relationship between excess cancer rates and exposures to cadmium compounds in these workers provides an estimate of lung-cancer potency of inhaled cadmium compounds. A simple linear regression provides a potency estimate of 14.6 x 10-6 per µg-yr/m3. A worker inhaling air at 10 m3/d, 5 d/wk, for 48 wk/yr inhales 10 x 5 x 48 = 2,400 m3/yr. Thus, the lung-cancer potency estimate is (14.6 x 10-6)/ 2,400 = 6.1 x 10-6 per milligram of cadmium. An exposure of a 70-kg person to ' mg over 75 yr amounts to a lifetime average daily exposure of 1/(70 x 365 x 75) = 0.52 x 10-6 mg/kg-day. The cancer-potency risk factor and exposure are both expressed on a body-weight basis. Here, this has no effect on the estimate of risk. The cancer risk estimate is based on the cumulative exposure of humans. Thus, there is no need for a cross-species dose-scaling factor based on, for example, lung surface area. The lung-cancer potency estimate expressed in terms of lifetime average daily exposure is (6.1 x 10 -6)/(0.52 x 10-6) =

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests 11.7 per milligram of cadmium per kilogram body weight per day. The estimate of cancer potency based on human data is almost twice the upper limit for potency of 6.3 per milligram per kilogram body weight per day of cadmium based on inhalation studies in rodents (Integrated Risk Information System, Office of Health and Environmental Assessment, EPA, Washington, D.C.). The studies in workers exposed to cadmium do not indicate which particular compounds are responsible for the carcinogenic response. The potency value for lung cancer might reflect "bioavailable" cadmium, but it might also reflect cancer caused by the specific cadmium compounds present. It is reassuring that the cancer potency estimates for cadmium based on animal or human data are within a factor of 2 of each other. The National Research Council's report Science and Judgment (NRC 1994) states that "the carcinogenic risk associated with specific cadmium compounds could be overestimated or underestimated, because bioavailability has not been included in the risk assessment." However, the Occupational Safety and Health Administration came to the conclusion (OSHA 1992): "Record evidence and expert opinion [given by G. Oberdörster and U. Heinrich] led the Agency to conclude that CdS should be considered an occupational carcinogen and assigned the same PEL [permissible exposure limit] as that established for other Cd compounds." In estimating the potential lung-cancer risk associated with exposure to ZnCdS, the subcommittee followed the procedure adopted by OSHA and assumed equal bioavailability of cadmium from all cadmium compounds. On the basis of the subcommittee's expert judgment and the data presented in Chapters 3 and 4, it is highly unlikely that this is the case for ZnCdS. LUNG-CANCER RISK ESTIMATES Releases of ZnCdS at all locations are shown in Table 5-1 and Appendix B, Table B-1. Maximal cumulative exposures reported at a site were expressed, in terms of ZnCdS, as microgram-minutes per cubic meter. Multiplying this by 0.0166 m3/min yields an inhaled dose of ZnCdS. Multiplying that by 0.156 yields an inhaled cadmium dose. The lifetime average

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests daily exposure was calculated by dividing by 27,375 days (365 x 75) and by 70 kg to convert to a body-weight basis. The average daily exposure expressed as milligrams of Cd per kilogram per day was multiplied by the human lung-cancer potency factor, 11.7, to obtain an estimate of the lung-cancer risk. The estimates were multiplied by 10 to account for possible increased sensitivity of children (Table 6-2). For example, the maximum potential exposure reported for Biltmore Beach, FL was 150,000 µg-min/m3 of ZnCdS (Appendix B, Table B-1). Using an inhalation rate of 1 m3/h, an active person inhales 1/60th of a cubic meter of air per minute, i.e., 1/60 = 0.0166 m3 per minute. The maximum potential exposure was 0.0166 x 150,000 ≈ 2,500 µg of ZnCdS. Because 15.6% of ZnCdS is cadmium, the maximum potential exposure to cadmium is 0.156 x 2,500 = 390 µg, as reported in Table 6-2. Averaged over a lifetime of 75 years x 365 days per year = 27,375 days, the average daily exposure is 390 divided by 27,375 = 0.0142 µg of cadmium per day. For a body weight of 70 kg (148 pounds), the average daily exposure per kg of body weight is 0.0142 divided by 70 = 0.0002 µg/kg of cadmium. The human lung-cancer potency factor (probability of lung cancer per mg/kg cadmium per day average lifetime exposure) calculated earlier in this chapter is 11.7 per mg/kg of cadmium per day over a lifetime. Since one µg is 1/1,000th of a mg, the potency factor is 11.7 divided by 1,000 = 0.0117 per µg/kg of cadmium per day. The lifetime risk (probability of lung cancer) is estimated to be less than the potency factor times the average daily lifetime exposure: 0.0117 x 0.000204 = 0.0000024 = 2.4 x 10-6 (2.4 per million). The estimates were multiplied by 10 to account for possible increased sensitivity of children, 10 x 2.4 x 10-6 = 24 x 10-6 (24 per million) as reported in Table 6-2. For this particular remote beach area, it is unlikely that anyone was exposed at this level. The highest cancer risk estimates are at Biltmore Beach in the vicinity of Panama City, FL. Air samples there were collected on a remote beach over an area several yards wide in the center of the plume. It is unlikely that many people, if any, were exposed to such doses.

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests TABLE 6-2 Upper-Bound Estimates of Cumulative Airborne Cadmium Exposure and Lung-Cancer Risk Associated with Exposures of Potentially Sensitive Populations (Such as Children) at Locations with Maximal Recorded Exposurea Exposure Site Estimated Maximal Exposure per Person, µg Estimated Maximal Lifetime Riskc Biltmore Beach, FLb 390 24.0 x 10-6 St. Louis, MO (1963-1965) 19.2 1.2 x 10-6 Winnipeg, MB, Canada 14.5 0.9 x 10-6 Minneapolis, MN 6.8 0.4 x 10-6 St. Louis, MO (1953) 5.2 0.3 x 10-6 San Francisco, CA (1964-1967) 4.9 0.3 x 10-6 Palo Alto, CA (1962) 4.4 0.3 x 10-6 Chippewa National Forest, MN 4.2 0.3 x 10-6 Oceanside, CA (on shore) 1.8 0.1 x 10-6 San Francisco, CA (1950) 1.1 0.07 x 10-6 Ft. Wayne, IN 1.1 0.07 x 10-6 Dugway, UT 1.1 0.07 x 10-6 Clinton School, Minneapolis 0.5 0.05 x 10-6 Camp Cooke, CA 0.4 0.03 x 10-6 Oceanside, CA (offshore) 0.4 0.03 x 10-6 NC, SC, GA 0.3 0.02 x 10-6 Camp Detrick, MD 0.2 0.01 x 10-6 Greenbrier Swamp, MD 0.1 0.01 x 10-6 Cushing, OK 0.1 <0.01 x 10-6 Corpus Christi, TX <0.1 <0.01 x 10-6 Goldfield, NV <0.1 <0.01 x 10-6 Colfax, WA <0.1 <0.01 x 10-6 Stanford Univ., CA <0.1 <0.01 x 10-6 Palo Alto, CA (1950) <0.1 <0.01 x 10-6 a Almost all the people had exposures (and risks) lower by a factor of 2-10. Calculations assumed 100% bioavailability of cadmium from ZnCdS. b It is unlikely that many people, if any, were exposed to such doses because it was a remote, unpopulated beach area. c Twenty-four additional cancers in 1 million persons.

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests The highest estimates of lung cancer risk for a populated area were for St. Louis, MO. The Army conducted tests in St. Louis in 1953 and again during 1963-1965. It is possible that some individuals were in the maximally exposed areas during both tests. Their total maximum exposure could have been (19.2 + 5.2) = 24.4 µg of cadmium with an estimated lifetime lung cancer risk of less than (1.2 + 0.3) x 10-6 = 1.5 x 10-6. Examination of records and isopleth maps of the levels of ZnCdS measured in the air indicate that the maximum exposures occurred over a very small fraction of the test areas. For rooftop and street-level releases of ZnCdS, the vast majority of the population were exposed at less than 1/10th the maximum recorded values. Hence, the estimated lung-cancer risks for the vast majority of these populations are less than 1/10th that of the maximally exposed individuals as reported in Table 6-2. For airplane releases of ZnCdS, the vast majority of the population were exposed to less than one-half the maximum recorded values. Hence, their estimated lung-cancer risks are half (or less than half) that reported in Table 6-2 for the maximally exposed individuals. UNCERTAINTY OF LUNG-CANCER RISK ESTIMATES The major assumptions made and their effects on cancer risk estimates are listed in Table 6-1. It was assumed that all the cadmium in the ZnCdS particles was bioavailable. If only a fraction, such as 50%, of the cadmium were bioavailable, the risk estimates would be half of those calculated. The lung-cancer potency was estimated from results in workers exposed to cadmium compounds at higher air concentrations for periods much longer than those of any people exposed in the ZnCdS releases. It is unknown whether the cancer potency (risk) is lower for shorter periods. The values chosen for the breathing rate of active people (1 m3/h) and life span (75 yr) have little influence on cancer risk estimates. Undoubtedly, some people were exposed to cadmium concentrations somewhat higher than those recorded at any air sampler location. At Minneapolis and Ft. Wayne, where cadmium air concentrations were estimated from isopleths, it is unlikely that exposure could have exceeded the reported values by a factor of 10. In fact, lung cancer risk estimates

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests are unlikely to be low by a factor of 10 due to underestimates of air concentrations for the maximum exposed individuals. OTHER ROUTES OF EXPOSURE The cancer risk estimates in this section were based on direct exposure to airborne ZnCdS inhaled during the test releases. Some ZnCdS from the test releases would also settle on the ground and result in exposure via inhalation of recirculated dust particles. The workers on whom the cancer potency estimate was based were also exposed to recirculated dust particles. Hence, the contribution of this route of exposure, if any, to lung cancer was included in the potency estimate. CONCLUSIONS In the absence of adequate toxicity data on ZnCdS, the subcommittee considered it to be most appropriate to base its assessment of the potential toxicity of ZnCdS for noncancer health effects on the toxicity of CdS, an insoluble cadmium compound. Results of several toxicity studies conducted with CdS indicate that CdS inhaled into the lung is not solubilized and absorbed into the blood for distribution to other organs. Therefore, the subcommittee concluded that sporadic exposures of the public to small amounts of ZnCdS should not result in any noncancer systemic toxicity (toxicity to the kidneys, bone, immune system, or reproductive and developmental system). The only organ considered to be potentially at risk because of exposure to particles of ZnCdS was the respiratory tract; the particles of ZnCdS were respirable and would be expected to deposit deep in the lungs. Interaction of lung tissue with inhaled particulate matter might create circumstances that would assist in metal mobilization, and the release rate could be slow. Results of several toxicity studies indicate that CdS has little toxicity in the respiratory tract. Rats exposed to CdS at 39,600 mg-min/m3 had only a mild pulmonary response. Using that result and dividing by 10 to extrapolate from a LOAEL to a NOAEL, by 10 to extrapolate from animal to human exposures, and by 10 to account for

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests sensitive populations, one would not expect adverse health effects, even in sensitive populations, from exposure to CdS in an insoluble form, such as ZnCdS, at 39.6 mg-min/m3 or 39,600 µg-min/m3 or inhaled cadmium dose of 513 µg. The inhalation dose of 513 µg was considered to be a safe level of exposure and was compared with the maximal cadmium exposures at the test sites. The four highest cadmium doses that could have been inhaled were estimated to be 6.8 µg (Minneapolis), 24.4 µg (St. Louis), 14.5 µg (Winnipeg), and 390 µg (Biltmore Beach, FL). The exposures to cadmium from ZnCdS did not exceed the level considered to be safe at any test site. On the basis of the above information, the subcommittee concludes that the exposures to ZnCdS from the Army's tests (for which data are available) should not have caused noncancer health effects in exposed persons. Cancer risk estimates are based on the cadmium content of ZnCdS because no studies on the carcinogenicity of ZnCdS have been conducted and cadmium is the most-toxic component. This approach will overestimate the risk. Upper-bound estimates of lung-cancer risk based on 100% bioavailability of the inhaled cadmium from ZnCdS particles at the maximal recorded exposure are summarized in Table 6-2. For each city or town, the location with the highest average reported air concentration of Cd was used to estimate cancer risk. The risk estimates were multiplied by a factor of 10 to account for exposure to potentially sensitive subpopulations, such as children. The subcommittee believes it is unlikely that lung-cancer risk estimates have been underestimated. The potential exposures of people with the highest ZnCdS exposures from the Army tests in St. Louis were equivalent to the ambient airborne cadmium exposures of people living in typical urban areas for 1-8 mo. For the vast majority of people exposed to ZnCdS from the Army tests in St. Louis, the exposures were equivalent to the ambient airborne exposures of people living in typical urban areas for 1-3 wk. The estimated upper-bound lung-cancer risks are small even for people with the larger exposures. On the basis of a description of the tests and, in some cases, isopleths of exposure, the vast majority of people were exposed at concentrations that were less than the highest by a factor of 2-10 or even more. Accordingly, their cancer risks would be lower by a factor of 2-10 or more. It is un-

OCR for page 67
Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests likely that any people have experienced total exposures to cadmium much above the exposures that would have been experienced by a person who was at the maximally exposed location for each of the ZnCdS tests in a particular locale. It is unlikely that exposure estimates have been underestimated for the vast majority of the exposed population. Because of the extremely low concentrations of ZnCdS (or cadmium from ZnCdS) in the air and the short duration of exposure, the lung cancer risk, if any, is very low. Hence, it is unlikely that anyone in the test areas developed lung cancer owing to direct exposure to cadmium from airborne releases of ZnCdS.