Drinking Water and Health,: Volume 6 (1986)

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

9 Toxicity of Selected Contaminants The 14 compounds reviewed in this chapter were evaluated at the request of the EPA to assist the agency in regulating contaminants in drinking water. In selecting compounds for review, the committee was guided both by EPA's regulatory agenda and by concerns about important current toxicological issues within the research community. The 14 substances selected were, in order of discussion, acrylamide, aldicarb, diallate, sul- fallate, dibromochloropropane, 1,2-dichloropropane, 1,2,3-trichloropro- pane, 1,3-dichloropropene, di(2-ethylhexyl) phthalate, mono(2-ethylhexyl) phthalate, ethylene dibromide, nitrofen, pentachlorophenol, and trichlor- fon. Ten of the 14 contaminants are reviewed by a Safe Drinking Water Committee for the first time. The other four compounds, which were discussed in previous volumes of Drinking Water and Health, are reeval- uated in this volume. Whenever possible, the committee evaluated pub- lished, peer-reviewed literature pertaining to the compounds under study. For trichlorfon and di(2-ethylhexyl) phthalate, however, it examined re- views prepared by the World Health Organization, followed up by tele- phone calls to the investigators or knowledgeable sponsors. For acrylamide, aldicarb, and nitrofen, important new information was made available to the committee by researchers with projects under way. The committee conducted its own peer review of the unpublished studies and in some cases subjected the data to independent review. At the first stage of evaluation, an intensive literature review was con- ducted for each substance. In addition, data summaries were obtained from several offices of EPA, including the Office of Drinking Water and the Office of Pesticides. These summaries were used as an initial indication 294

Toxicity of Selected Contaminants 295 of the range of available toxicological data. In some cases, foreign lit- erature was translated and evaluated. Much of the data are the results of 2-year chronic feeding studies in rodents, reflecting past interests in car- cinogenesis testing. However, the committee carefully examined toxico- logical data on other effects, such as teratogenesis, mutagenesis, reproductive effects, and metabolism. In addition, it reviewed the relatively sparse data on current production, manufacture, environmental distribution, and en- vironmental monitoring. The committee recognized that ingestion may not be the sole route of exposure to contaminants in drinking water. Cooking, showers, bathing, swimming, and other activities could theoretically provide important toxic contributions; however, given the absence of data on these noningestion routes, the committee declined to develop specific estimates of exposure for them. In addition, drinking water is not the only means of exposure to many of the compounds evaluated here. All sources of exposure must be considered by regulators in setting acceptable levels of exposure to contaminants in water, regardless of the biological end point under con- sideration. To allow for exposures through other routes, the committee generally assumed that drinking water provided 20% of the total exposure to a given compound. Following its review of the toxicology data, the committee classified compounds according to whether they were or were not known (or sus- pected3 carcinogens. For carcinogens, the risk to humans was expressed as the probability that persons weighing 70 kg would develop cancer some time in their lives as a consequence of ingesting 1 liter of water containing 1 fig of the substance daily over a lifetime of 70 years. Although risks to the 10-kg child were not calculated, the disproportionately high intake of drinking water by children as compared with that of adults would place them at greater risk. The committee then examined models for extrapolating from the high doses used in animal studies to the lower doses common in the environment of humans, and concurred with many experts who believe that several risk quantification techniques should be utilized to produce an estimated range of risks rather than a single number. The selection of models for low-dose extrapolations mus; be somewhat arbitrary except for the mul- tistage model, which is the only one with firm biological criteria at this time. Nonetheless, the committee recognizes that risk quantification re- mains an essential tool for rationalizing regulatory actions. The computation of risk depends on several factors, ranging from the selection of a mathematical model for low-dose extrapolation to the as- sumptions made within the model to fit a specific computer program. Because of the uncertainties involved in determining the true shapes of the dose-response curves used for extrapolation and because recent re-

296 DRINKING WATER AND H"LTH search indicates several stages in cancer induction, the committee decided that in general the multistage model is the most useful. It appears to have more of a biological basis than most other models and in most cases is more conservative, giving higher estimates of risk at low doses than most other models. The model incorporates the reasonable assumption of back- ground additivity and is thus linear at low doses. Within the multistage model, one can compute the dose associated with a given level of risk by using either a restricted model or a generalized model. In the restricted model, the number of possible stages in the multistage process is limited to the number of doses at which the exper- iment was conducted minus 1. The generalized model places no such limit on the possible number of stages; rather, it permits computation of the best fit to the data without this constraint. A more detailed discussion of the committee's reasoning in selecting the generalized multistage model and its overall framework for risk assessment appears in Chapter 8. Although, as stated in Chapter 8, the committee believes that an un- derstanding of pharmacokinetic principles is useful to the extrapolation of response at high doses to estimate response at low doses, the relative paucity of pharmacokinetic data is apparent in the risk assessments made in this chapter. Adequate data of this type were not available for any of the compounds studied. The committee recommends a review of the needs and potential gains possible through the use of pharmacokinetic data and, where appropriate, stimulation of the acquisition of such data for com- pounds under consideration for future risk assessments. For agents not identified as known or suspected carcinogens and for which there were adequate toxicity data from prolonged ingestion studies in humans or animals, the committee calculated an acceptable daily intake (ADI), using methods developed in earlier volumes of this series and estimating dose-response relationships when data were sufficient. This conventional approach was taken by default in the absence of suitable low-dose extrapolation models and because a "safe level" has not been demonstrated for these noncarcinogenic effects. For carcinogens that pro- duced other toxic effects at low levels, the committee also estimated the minimum exposure levels at which such effects might be expected to occur. The ADI is derived by estimating the no-observed-effect level (NOEL) for any given compound and then dividing it by an uncertainty or safety factor. Aware of the pitfalls encountered in estimating NOELs, the committee carefully weighed the evidence supporting this level in any given study. Also sensitive to possible misinterpretations concerning the use of "safety" factors, the committee recognized that these factors prop- erly indicate levels of confidence in the underlying studies. For some compounds, the data base was adequate to permit an estimate of the magnitude of inter- or intraspecies variability and suggest a safety factor

Toxicity of Selected Contaminants 297 based on that estimation. Where such an estimation was not possible, the committee used safety factors provided in previous guidelines: 10 when satisfactory data from chronic epidemiological or clinical studies were used; 100 for well-conducted long-term animal studies; and 1,000 for short-term studies or studies with some potential inadequacies. The poly- morphism of human drug metabolism indicates that the range of intra- human variability may be as high as 100-fold, implying that the uncertainty factor of 10 may not be adequately conservative. Furthermore, when extrapolating risk to the general population from epidemiological data, lack of quantitative exposure data may necessitate a further uncertainty factor. ACRYLAM I DE 2-Propenamide CAS No. 79-06-1 RTECS No. AS3325000 H O 1 11 CH2 = C—C—NH2 Acrylamide, the neurotoxic monomer of a commercially important poly- mer, polyacrylamide, is a highly reactive agent that spontaneously reacts with hydroxyl-, amino-, and sulfhydryl-containing compounds (Hashi- moto and Aldridge, 1970). It has a molecular weight of 71.08, a melting point of 84.5°C, and a vapor pressure of 0.007 mm mercury at 20°C. Its solubility in water is 215 g/100 ml at 30°C. Total U.S. production of acrylamide in 1978 was 34,000 metric tons (MacWilliams, 1978~. Acrylamide polymers are used as additives to enhance oil recovery, in- crease dry strength in paper products, dissipate fog, and stabilize soil. They are also used in grouting operations, clarification of potable water, and treatment of municipal and industrial effluents. The biodegradation and environmental fate of acrylamide have been examined in both water and soil. Cherry et al. (1956) found that acrylamide degraded in filtered river water in 10 to 12 days. However, Croll et al. (1974) found more rapid degradation in river water of approximately 4 days. When acrylamide was added to soil (Lance et al., 1979), complete degradation occurred in approximately 6 days; a maximum of 60% of the acrylamide was degraded to carbon dioxide. Acrylamide should not sig- nificantly accumulate in the environment because of its high water solu- bility (Dow Chemical USA, 19841.

298 DRINKING WATER AND HEALTH METABOLISM Following administration, acrylamide is rapidly distributed to all tissues, metabolized, and excreted (Edwards, 1975; Miller et al., 19821. After a single dose of ~4C-labeled 2,3-acrylamide, Miller et al. (1982) found equivalent concentrations of acrylamide in all tissues except in erythro- cytes, where acrylamide appears to accumulate (Pastoor and Richardson, 19811. Nervous system tissues accumulated less than 1% of the dose of acrylamide. The tissue content of radiolabeled acrylamide decayed in biexponential fashion (half-life of approximately 8 days), except in eryth- rocytes, where shortly after dosing a plateau was reached with a half-life of approximately 10.5 days (Pastoor and Richardson, 19811. The major route of biotransformation of acrylamide is conjugation with the tripeptide glutathione (Miller et al., 1982; Pastoor et al., 1980~; it is eventually excreted in the urine as N-acetyl-S-~3-amino-3-oxypropyl~cysteine. This route appears to be detoxifying since depletion of the nonprotein sulfhydryl content increases the neurotoxic potency of acrylamide. In addition to conjugation with glutathione, acrylamide appears to un- dergo partial microsomal-mediated metabolism. HEALTH ASPECTS Observations in Humans Although most human exposures to acrylamide result from dermal ab- sorption or ingestion of dust, one report documented exposure and toxicity resulting from drinking water contaminated with acrylamide. Igisu et al. (1975) described a Japanese family of five who ingested well water con- taminated with acrylamide from a nearby grouting operation. The con- centration of acrylamide in drinking water was found to be approximately 400 mg/liter. The children, who consumed acrylamide-free water while at school, developed mild gait disorders and sleep disorders. The parents, who consumed the ac~lamide-contaminated water exclusively, developed slurred speech, unsteady gait, memory loss, irrational behavior, and vi- sual, tactile, and auditory hallucinations. In other studies in humans, all the investigators (Auld and Bedwell, 1967; Davenport et al., 1976; Fullerton, 1969; Garland and Patterson, 1967) described neuropathy with a consistent set of symptoms. The first clinical manifestation of acrylamide neuropathy in humans is slowly pro- gressing symmetric distal sensory abnormalities and motor weakness. Sub- jects commonly reported skin sensitization (contact dermatitis); cold, blue hands; unsteadiness; muscle weakness; paresthesia; and numbness of the

Toxicity of Selected Contaminants 299 hands or feet. Tendon reflexes disappear and vibrational sense is lost, but heat, pressure, and other objective sensory modalities remain intact. Re- covery from mild forms of acrylamide neuropathy is usually complete, occurring within a few months. Patients with severe neuropathy may never completely recover, but may experience residual ataxia, distal weakness, and sensory loss. Observations in Other Species Acute Effects The oral LD50 for acrylamide is 150 to 180 mg/kg body weight (bw) in rats, guinea pigs, and rabbits and 100 mg/kg bw in cats and monkeys (McCollister et al., 19641. Symptoms of intoxication in cats after high doses of acrylamide include behavioral disturbances, clonic seizures, severe ataxia, tremors, and death due to respiratory failure (Kuperman, 19581. Similiar responses are reported for rats, guinea pigs, rabbits (McCollister et al., 1964), and chickens (Edwards, 1975~. Miller et al. (1983) demonstrated inhibition of retrograde axonal transport in peripheral nerves within 24 hours after intraperitoneal administration of acrylamide to rats at 40 mg/kg. Increased numbers of dopamine receptors have been reported in the corpus striatum of rats following oral treatment with a 25-mg/kg bw dose of acrylamide (Agrawal et al., 19811. Subacute and Chronic Elects Subacute and chronic exposures to acrylamide have been demonstrated to produce symptoms of peripheral neuropathy in cats, rats, mice, guinea pigs, rabbits, and monkeys (McCollister et al., 19641. As with humans, intoxicated animals develop limb incoordination, which progresses to ataxia and weakness. This is most obvious in the hind limbs. Additional symptoms in acrylamide- intoxicated animals include weight loss, enlarged and distended bladders, and testicular atrophy. The enlarged bladders were attributed to "nervous retention"; however, the animals did continue to pass urine. The duration of exposure to acrylamide required to produce neuropathy appears to be a direct function of the magnitude of the acrylamide dose. Kuperman (1958) demonstrated that cats treated with acrylamide (1 to 50 mg/kg/day) developed clinical signs of acrylamide intoxication after re- ceiving an average total dose of approximately 102 mg/kg by a variety of routes, including intravenous and intraperitoneal administration, indepen- dent of whether the dose was administered over 2 days or 4 months. These data indicate that acrylamide is a cumulative neurotoxicant. Fullerton and Barnes (1966) demonstrated that degeneration of the distal processes of large-diameter peripheral nerves was associated with acrylamide-induced neuropathy in rats. Exposure to acrylamide in the diet at daily doses of approximately 15 to 18 mg/kg bw for 10 weeks resulted

300 DRINKING WATER AND HEALTH in severe axon loss and proliferation of Schwann's cells in distal peripheral nerves. Pathology was most evident in the longest fiber tracts containing large-diameter axons. Electron microscopic studies of acrylamide-treated rats conducted by Prineas (1969) revealed dramatic increases in the number of axoplasmic neurofilaments and organelles. Occasional axons demon- strated invaginations of finger-like Schwann's cell processes that may act to remove damaged or degenerating axonal constituents. The dose of acrylamide required to produce neuropathy following chronic exposure has been investigated in several species. Structural or functional necrologic deficits have been noted after daily oral administration of acrylamide to rats (1 mg/kg bw for 93 days) (Burek et al., 1980), cats (0.7 mg/kg bw for 240 days) (McCollister et al., 1964), and monkeys (1 mg/kg bw for 18 months) (Schaumburg et al., 19821. Mutagenicity Acrylamide has been reported to be nonmutagenic in Salmonella typhimurium strains TA1535, TA1537, TA98, and TA100 at doses of 0.001 to 3 mg/plate with and without metabolic activation (Bull et al., 19841. Lack of acrylamide-induced genotoxicity was confirmed in the hepatocyte primary culture DNA repair test (Miller et al., 19841. Carcinogenicity Bull et al. ~ 1984) investigated the carcinogenic effects of acrylamide by administering it to female Sencar mice six times over 2 weeks at oral and intraperitoneal doses of 12.5, 25, and 50 mg/kg. The shaved back of each animal was subsequently treated with 1 1lg of 12-O- tetradecanoyl-phorbol-13-acetate (TPA) three times a week for 20 weeks, and the animals were sacrificed after 52 weeks. Acrylamide was found to produce dose-dependent increases in incidence of squamous cell car- cinoma. In addition, it produced dose-dependent decreases in the time to tumor appearances. No tumors were seen in animals not treated with TPA. In a separate experiment, Bull et al. (1984) found that oral or intraperi- toneal administration of acrylamide to A/J mice at doses of 6.25, 12.5, and 25 mg/kg three times a week for 8 weeks resulted in dose-dependent increases in the incidence of lung adenomas when measured 4 months after the last dose of acrylamide was given. K. A. Johnson et al. (1984) reported that male and female Fischer 344 rats developed tumors following a 2-year exposure to acrylamide in drink- ing water. Male rats exposed to ac~lamide at 0.5 mg/kg bw a day for 2 years developed scrotal mesotheliomas. At 2.0 mg/kg/day, benign thyroid tumors were also observed in male rats, whereas females demonstrated benign and malignant thyroid tumors, glial tumors within the central ner- vous system, adenomas of the clitoral gland, squamous cell papillomas in the mouth, benign and malignant mammary tumors, and malignant uterine tumors.

Toxicity of Selected Contaminants 30 ~ TABLE 9-1 Tumor Incidence in Rats Fed Acrylamide-Contaminated Drinking Watera Animal Tumor Dose Sex Site (mglkg/day) Tumor Rates 3/60 0/60 7/60 1 1/60 10/60 Fischer 344 rat Male Scrotum O 0.01 0.1 0.5 2.0 aBased on data from K. A. Johnson et al., 1984. Carcinogenic Risk Estimate In the drinking water study recently com- pleted by K. A. Johnson et al. (1984), there was an increased incidence of scrotal mesothelioma in male Fischer 344 rats. In the study by Bull et al. (1984), there was an increase in lung tumors. The tumor incidences from the K. A. Johnson et al. (1984) study and the Bull et al. (1984) study are summarized in Table 9-1 and Table 9-2, respectively. Using these data, the committee estimated the lifetime risk and upper 95% confidence estimate of lifetime risk in humans after a daily con- sumption of 1 liter of water containing acrylamide at a concentration of 1 ,ug/liter. The conversion of animal to human doses is based on body surface area, assuming the following weights: humans, 70 kg; rats, 400 g; and mice, 33 g. The conversion formula is: animal consumption = human consumption times (human weight/animal weight)i'3. The risk estimates calculated with the generalized multistage model are shown in Table 9-3, and those based on the Weibull model are shown in Table 9-4. It is useful to compare the results obtained from the generalized multistage model with those of the Weibull model, which appeared to fit the data better. (See the discussion of risk-assessment models presented in Chapter 8.) TABLE 9-2 Tumor Incidence in Mice Given Acrylamide Intraperitoneallya Animal Tumor Site Dose (mg/kg/day) Tumor Rates Males 2/16 8/16 6/16 10/17 14/15 Females 1/15 6/17 9/17 1 1/14 14/15 A/J mouse Lung o 3 10 30 aBased on data from Bull et al., 1984.

302 DRINKING WATER AND HEALTH TABLE 9-3 Carcinogenic Risk Estimates for Acrylamide from the Generalized Multistage Modela Upper 95% Confidence Estimated Human Estimate of Lifetime Animal Sex Lifetime Riskb Cancer Riskb Fischer 344 rats Male 6.6 x 10-6 1.2 x 10-5 A/J moused Male 3.8 x 10-6 7.5 x 10-6 A/J moused Female 8.2 x 10-6 1.4 x 10-s aFrom GLOBAL83, a software program developed in 1983 by R. B. Howe and K. S. Crump; modified for microcomputer compilation in 1985 by M. S. Cohn, U.S. Consumer Product Safety Commission, Washington, D.C. bAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ~g/liter. CBased on data from K. A. Johnson et al., 1984. Based on data from Bull et al., 1984. In previous volumes of Drinking Water and Health, the risk estimates for male and female rats and mice were averaged to yield one composite number. If the data for the generalized multistage model in Table 9-3 are averaged, the estimated human lifetime risk is 1.2 x 10-s; the upper 95% confidence estimate of lifetime cancer risk is 2.1 x 1o-s . Developmental Effects Treatment of rats with acrylamide at 200 or 400 ppm daily (20 to 40 mg/kg bw) during gestation has been shown to produce no gross or histologic evidence of teratogenicity (Edwards, 1976~. CONCLUSIONS AND RECOMMENDATIONS Acrylamide is a highly reactive molecule that produces peripheral neuropathy in animals and humans following repeated exposure. The mag- TABLE 9-4 Carcinogenic Risk Estimates for Acrylamide from the Weibull Model Upper 95% Confidence Estimated Human Estimate of Lifetime Animal Sex Lifetime Riska Cancer Riska Fischer 344 ratb Male 1.7 x 10-4 9.7 x 10-4 A/J mouser Male 4.0 x 10-5 1.5 x 10-4 A/J mouser Female 7.4 x 10-5 2.7 x 10-4 aAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ~g/liter. bBased on data from K. A. Johnson et al., 1984. CBased on data from Bull et al., 1984.

Toxicity of Selected Contaminants 303 nitude of the dose of acrylamide required to produce neuropathy is in- versely related to the duration of exposure. Thus, no-observed-effect levels determined from laboratory studies of relatively short duration (less than 2 years) may be of little value in determining human risk following lifetime exposure. For this reason and because acrylamide is a cumulative toxicant, no further risk assessment was attempted for neurotoxicity. Recent data demonstrate that acrylamide is carcinogenic in laboratory animals. The estimated lifetime risk and the upper 95% confidence esti- mate of lifetime risk of cancer in humans presented above are based on both the multistage and Weibull low-dose extrapolation models. CARBAMATE PESTICIDES (ALDICARB, DIALLATE, AND SULFALLATE) ALDICARB 2-MethyI-2-(methy~thio~propanal ~~(methylamino~carbonyI]oxime CAS No. 6-06-3 RTECS No. UE2275000 CH2 0 CH3S—C—CH = NOCNHCH3 CH2 Since aldicarb was reviewed in Volumes 1 and 5 of Drinking Water and Health (NRC, 1977, pp. 635-643; 1983, pp. 10-12), the following section is primarily an examination of data not considered by the previous committees. HEALTH ASPECTS Observations in Humans The committee subjected to peer review a project report (Cope and Romine, 1973) on acute oral exposure of 12 healthy male volunteers (four males per group). This report indicated that aldicarb doses of 0.025 ma/ kg bw produced approximately 50% inhibition of blood cholinesterase, as measured by a radiometric technique. Cholinesterases are a family of enzymes responsible for hydrolyzing esters of choline such as acelylcho- line or butyrylcholine. At the highest dose, 0.1 mg/kg bw, approximately 70% inhibition of blood cholinesterase occurred attended by signs and

304 DRINKING WATER AND HEALTH symptoms of hypercholinergic action. The signs and symptoms of poi- soning were for the most part gone within 4 hours after dosing, and the blood cholinesterase level was normal after 6 hours. Observations in Other Species Acute Effects Cambon et al. (1979) gave corn oil solutions of aldicarb to pregnant female Sprague-Dawley rats by gastric intubation (eight rats per group fasted 24 hours). The doses were 0.1, 0.01, and 0.001 mg/kg bw. The effect of this treatment on fetuses was discussed in Volume 5 of Drinking Water and Health (NRC, 1983, pp. 10-12), but the significant inhibition of maternal erythrocyte and liver plasma cholinesterase in dams given 0.1 and 0.01 mg/kg bw was not noted. Subacute Elects Aldicarb sulfoxide is a more effective esterase in- hibitor in vitro than the parent compound, whereas aldicarb sulfone is a poor esterase inhibitor. DePass et al. (1982) gave a 1:1 mixture of aldicarb sulfoxide/aldicarb sulfone to Wistar strain rats (10 rats of each sex per dose level range) in their drinking water for 29 days. The authors reported depressed body weight and food consumption 7 days into the exposure period in a group given 19.2-ppm concentrations of the mixture.These parameters remained depressed throughout the study in males (statistically analyzed on days 14, 21, and 29) but returned to normal in females. However, erythrocyte and plasma cholinesterase activity remained de- pressed after 8, 15, and 29 days in both males and females exposed to 19.2 ppm. Statistically significant reduction of plasma cholinesterase was seen in males exposed to 4.8 ppm for 8 days but not on later assay dates; erythrocyte cholinesterase was not depressed in this exposure group until day 29. Other exposure levels, where effects were not seen, were 1.2, 0.3, and 0.075 ppm. These data are difficult to interpret in that a mixture of aldicarb metabolites was used, and individual analyses of the two metabolites were not described. Chemical analysis revealed concentrations of the compound at only 80% of nominal values. Weil and Carpenter (1968) fed Harlan-Wistar rats (15 animals of each sex per group) aldicarb sulfoxide in the diet in concentrations that were believed to provide daily doses of 1.0, 0.5, 0.25, and 0.125 mg/kg bw. It does not appear that true doses were established. Five rats of each sex were sacrificed after 3 months for assay of plasma, erythrocyte, and brain cholinesterase using a titrimetric method. The surviving animals were killed after 6 months on the diet, and the same measurements were taken (except that a colorimetric assay was used). It was found that plasma and erythrocyte cholinesterases were generally more sensitive than brain cho- linesterases to aldicarb sulfoxide inhibition. The effect at the lowest dose

Toxicity of Selected Contaminants 305 was significant inhibition of plasma cholinesterase in males given daily diets providing 0.125 mg/kg bw for 3 months. Six months after treatment ended, this effect was not detected. Erythrocyte cholinesterase was in- hibited significantly in both sexes at doses of 0.25 mg/kg/day for 6 months, but effects of 0.125 mg/kg/day for this length of time are unclear. Brain cholinesterase was inhibited significantly by 1.0 mg/kg/day in both sexes at both exposure lengths; in females, it was inhibited by 0.5 mg/kg/day at 6 months. In a similar study, Well and Carpenter (1968) noted significant plasma cholinesterase inhibition in male rats (five per group) given diets providing aldicarb sulfoxide at 0.25 but not at 0. 125 mg/kg bw a day for 3 months. In both males and females, erythrocyte cholinesterase was significantly inhibited after 3 months on diets providing 0.5 mg/kg/day, but no depres- sion was observed at 0.25 mg/kg/day. In animals fed comparable diets but given a 1-day unadulterated diet prior to sacrifice, no inhibition of esterases was seen, suggesting that rapid reversal of inhibition occurs even after subacute exposures. Well and Carpenter (1968) also fed beagle dogs aldicarb sulfoxide diets providing 0.5, 0.25, 0.125, or 0.0625 mg/kg bw each day for 3 months. The only significant depression was found in plasma cholinesterase after 1 month on 0.5 mg/kg/day. These animals lost weight in the first week of the study but then returned to normal for the remainder of the experiment. Chronic Effects No new data were found by the committee. Mutagenicity There have been no published evaluations of the geno- toxicity of aldicarb. The National Toxicology Program has listed National Institute for Environmental Health Studies contracts to study the mutagenic and genotoxic effects of the compound in microbial systems and in Dro- sophila (NTP, 1984b). Carcinogenicity No new data were found by the committee. Developmental Elects No new data were found by the committee. CONCLUSIONS AND RECOMMENDATIONS Signs of hypercholinergic activity, such as lacrimation, salivation, meiosis, and convulsions, are unequivocally toxic effects of aldicarb. These effects result from inhibition of acetylcholinesterase at specific neuroeffector sites, and it appears that a substantial inhibition of enzyme activity is required before an effect is observed. Studies with organophosphorus anticholin- esterases suggest that greater than 50% inhibition of plasma acetylcholin-

306 DRINKING WATER AND HEALTH esterase would be required to produce overt hypercholinergic signs (Wills, 19721. Effects of low-level inhibition of acetylcholinesterase by aldicarb have not been encountered, although behavioral effects of long-term, low- level exposure to organophosphorus anticholinesterases have been sug- gested in the clinical literature (NRC, 1982a). Research on the unconfirmed effects of low-level inhibition of acetyl- cholinesterase or other esterases is to be encouraged. Although aldicarb inhibits other esterases, the physiological functions for these enzymes are not known. For example, serum cholinesterase is aldicarb sensitive, but the genetically determined absence of this enzyme does not appear to affect one's health (Kalow, 1965~. Serum cholinesterase inactivates the neuromuscular blocking agent succinylcholine (Taylor, 1980~. Thus, in- hibition of esterases by aldicarb might alter the metabolism of other xenobiotic substances. Orally administered aldicarb inhibits esterases in both animals and hu- mans at relatively low doses, compared to other carbamates. The onset of action and recovery from inhibition are rapid. In view of the transience of esterase inhibition, studies with drinking water or feed as a route of exposure probably lead to serious underestimations of the inhibition that would result from an equivalent dose given as a bolus. This may be part of the reason that doses considered NOELs in rodents (0.1-0.125 mg/kg bw over 24 hours) produced overt signs of hypercholinergic action when given to humans in a single, acute dose (0.1 mg/kg bw) (Cope and Romine, 19731. It appears that exposure through bolus administration is the worst- case circumstance that must be addressed when determining appropriate measures to avoid aldicarb toxicity. Unfortunately, data on esterase inhibition after bolus administration are limited. Cambon et al. (1979) have relevant data, but the assays were done on tissue from pregnant female rats and the appropriateness of that model is highly debatable. Cope and Romine (1973) obtained data from a very small number of healthy human males but did not estimate a NOEL. However, there are several considerations in favor of using their data: · The data pertain to healthy human male subjects, eliminating the uncertainty of interspecies extrapolation. · The dose was administered in drinking water. Thus, the route of exposure is identical to accidental drinking water ingestions. · Whole blood esterase was measured. This esterase activity is due primarily to red blood cell (RBC) acetylcholinesterase (Vandekar, 19801. Data from animal studies indicate that RBC acetylcholinesterase may be more sensitive to inhibition by oral aldicarb exposure than the acetylcho- linesterase of the relevant neuroeffector sites. Although RBC acetylcho- linesterase has no known physiological function, its use provides an additional

Toxicity of Selected Contam tenants 307 safety factor in determining daily intakes that would not be expected to produce toxicity. Wilkinson et al. (1983) have suggested, however, that the extreme dilution of central nervous system (CNS) tissue homogenates required for acetylcholinesterase assay may promote reversal of inhibition in vitro. This would result in underestimation of in viva CNS acetylcho- linesterase inhibition and provide an alternative explanation for the dis- parity in measurements between the CNS and erythrocytes. In addition, this small study in humans produced a dose-effect curve that could be used to extrapolate to doses of aldicarb expected to cause minimal inhibition of whole blood esterase, which is unlikely to be as- sociated with toxic effects. Because of the large variability in cholines- terase inhibition, a 20% inhibition of whole blood esterase was set as the point for extrapolation. This is a value that Gage (1967) suggested as the minimum to be considered abnormal, given the population variance in erythrocyte cholinesterase. Gage (1967) used a formula suggested by Cal- laway et al. (1951) and assumed that variation in erythrocyte cholinesterase activity in one person would be approximately 10%. Using these as- sumptions, he calculated that minimal significant inhibition would be 16.5% if derived from a mean of multiple preexposure activity determi- nations from one person, but would rise to 23% if based on a single preexposure determination. A previous Safe Drinking Water Committee discussed the use of 20% inhibition in estimating the ADI for various pesticides (NRC, 1980, p. 311. Many organizations have set 30% inhibition of red cell cholinesterase as a level at which action to prevent further exposure to organophosphorus anticholinesterases should be taken (Permanent Commission and Inter- national Association of Occupational Health, Subcommittee on Pesticides, 1972; Vandekar, 19801. In their evaluation of aldicarb residues in Long Island groundwater, Wilkinson et al. (1983) expressed the opinion that up to 40% inhibition of whole blood cholinesterase would not be likely to produce overt toxic effects. The maximum permissible cholinesterase inhibition in erythrocytes for repeated or continuous exposure before significant health effects occur cannot be determined from currently available data. The relationship be- tween erythrocyte cholinesterase activity and signs of toxicity varies with different substances and also may vary among individuals, depending on factors such as rate of distribution and metabolism of inhibitors (Murphy, 1980~. In this committee's view, repeated levels of 20% to 30% inhibition are important indicators of changes in biological status. Therefore, the committee decided to develop estimates for both these levels. The ex- trapolation was done on the upper 95% confidence interval of the dose. (The small number of subjects and the inherent variability in the population

308 DRINKING WATER AND HEATH TABLE 9-5 Estimated Dose of Aldicarb for Cholinesterase Inhibition from the Probit and Logit Dose-Response Models Dose Estimated by Dose Estimated by Percent Probit Model (mg/kg/day) Logit Model (mg/kg/day) Inhibition MLEa 95% UCLb MLEa 95% UCLb 20 0.0052 0.0001 1 0.0051 0.000105 30 0.0100 0.00052 0.0100 0.00051 aMaximum likelihood estimate. bUpper confidence limit. were, of course, reflected in the size of the confidence interval.) The committee also used the maximum likelihood estimate dose (i.e., the dose above which approximately 50% of the population would be affected) with an added safety factor of 10, because the test population consisted exclusively of healthy males. Both the 20% and 30% cholinesterase in- hibition in red blood cells were used to measure response against the logarithm of dose with the logit and the probit dose-response models. These mathematical models, more fully described in Chapter 8, both produce S-shaped dose-response cubes and give essentially the same dose- response relationship for aldicarb, even though they represent different statistical treatments. The results of the curve-fitting are given in Table 9-5. Because the percentages of inhibition data essentially reflect a contin- uous measure and are not the quantal response (yes-no) data that are usually fitted by tolerance (or individual sensitivity) distributions, the use of these models may be questionable for these data. However, the molecular interactions between chemical and enzyme may be considered quantal interactions for these types of models. Nonetheless, an alternative to the tolerance distribution models would be to fit a simple least-squares line to them, but such a dose-response curve has the inherent potential of estimating a less-than-zero or more-than-100% response at doses outside the observed range. Neither the probit model nor the logit model permits such antiintuitive estimates, and both reflect the S-shaped nature of the usual dose-response curve. These "reasonable" features of the models argue for their use in this context. The results of long-term feeding studies in rats and dogs have been reviewed in previous volumes of Drinking Water and Health (NRC, 1977, pp. 635-693; 1983, pp. 10-121. These studies have established no- observed-effect levels of 0.1 mg/kg bw a day in both species. Using a safety factor of 100 for the animal study results, the committee obtained a value (0.001 mg/kg/day) identical to that obtained in human studies with

Toxicity of Selected Contaminants 309 TABLE 9-6 Estimated ADIs for Aldicarb Based on 20% to 30% Cholinesterase Inhibition for Adults and Children Using Both the MLEa and the 95% UCLb Cholinesterase Adult (~g/liter) Child (1lg/liter) Inhibition (%) MLEa 95% UCLb MLEa 95% UCLb 20 3.5 0.7 1 0.2 30 7 3.5 2 1 aMaximum likelihood estimate from the probit and logit models. bUpper confidence limit. both the probit and logit dose-response models for 30% inhibition and a safety factor of 10. This value results in the calculation of a 0.007-mg/ liter ADI for aldicarb (NRC, 1983, p. 12~. Using the values obtained in Table 9-5, the committee calculated ADIs for both adults and children. For these calculations, it is assumed that a 70-kg adult consumes 2 liters of water daily, that a 10-kg child consumes 1 liter of water daily, and that 20% of the intake is provided by water. The following is an example of the calculation for an adult using the maximum likelihood estimate (MLE) from either the probit or logit dose- response models for 30% inhibition and an uncertainty factor of 10: 0.01 mg/kg low/day x 70 kg bw x 0.2 0.007 mg/liter, or 7 ~g/liter. 2 liters x 10 The results of these calculations are shown in Table 9-6. DIALLATE DiisopropyIthiocarbamic acid S-~2,3-dichIoroallyI) ester bis(~- methyl et try! ~ca rba moth ioate CAS No. 2303-16-4 RTECS No. EZB225000 CH3 CH3—CH CH3—CH CH3 0 C! C1 11 1 1 N—C—S—CH2—C = CH Diallate is an herbicide used for preemergence control of wild oats and other weeds in fields of sugar beets, flax, barley, corn, and a variety of

310 DRINKING WATER AND HEALTH other crops. It has a molecular weight of 270.2 and is a liquid at tem- peratures greater than 25°C to 30°C but less than 150°C. At 25°C, diallate has a vapor pressure of 0.00015 mm mercury and a density of 1.188 g/ml at 25°C. The herbicide is supplied in the United States as an emulsified concentrate containing 45% active ingredient and also as a powder consisting of 10% diallate. Diallate has a half-life in soil of approximately 30 days when applied at recommended levels of 1.5 to 3.5 pounds/acre (1.3 kg to 3.1 kg/ha). Estimated exposure to diallate during the application process is approximately 0.5 ~g/kg bw as a result of inhalation and 980 ~g/kg bw from dermal deposition (Dubelman et al., 19821. METABOLISM Data are not available. HEALTH ASPECTS Observations in Humans Data are not available. Observations in Other Species Acute Effects Oral administration of diallate to rats (500 to 1,600 ma/ kg bw) or guinea pigs (330 to 1,125 mg/kg bw) produced excitement and restlessness within 2 hours. There is increased sensitivity to pain and touch, ataxia, and hind limb paralysis. Animals died from respiratory failure within 1 to 4 days (Doloshitskii, 19691. The acute LDso was ap- proximately 1 g/kg bw in rats and 420 mg/kg bw in guinea pigs. Subchronic Effects Rats treated with oral diallate doses of 200 ma/ kg/day died within 18 days. Administration of 100 mg/kg/day resulted in the death of 76% of the animals tested. Prior to their death, white-blood- cell counts and neutrophils were significantly increased, lymphocytes were decreased, eosinophils were absent, coagulation time was increased, and weight loss was apparent (Doloshitskii, 19691. In another study (Palmer et al., 1972), sheep developed muscle spasms following three daily oral doses of diallate (25 mg/kg bw). Chronic E~ects Oral administration of diallate to rats at daily doses of 5, 20, and 50 mg/kg bw for 8 months produced increased blood sugar levels, inability to maintain body temperature, decreased serum cholin-

Toxicity of Selected Contaminants 3 ~ ~ TABLE 9-7 Hepatoma Incidence in Mice Exposed to Diallate (Avadex~a Tumor Rates Controls Treated Mouse Strain and Sex All Hepatic All Hepatic (C57BL/6 x C3H/Anf)F1 Males 22/79 8/79 14/16 13/16 Females 8/87 0/87 5/16 3/16 (C57BL/6 x AKR)F1 Males 16/90 5/90 12/18 10/18 Females 7/82 1/82 2/15 1/15 aBased on data from Innes et al., 1969. esterase activity, and alterations in central nervous system function (e.g., of rheobase and chronaxy). The apparent no-observed-effect level was 0.5 mg/kg bw a day (Doloshitskii, 19691. Neurotoxicity In cats, rats, and mice, diallate produced central nervous system excitation that rapidly progressed to clonic convulsions (Pestova, 19661. Diallate has also been reported to produce delayed peripheral neu- ropathy in hens (EPA, 19774. Mutagenicity Diallate produced base-pair mutations in TA100 and TA1535 strains of Salmonella typhimurium at doses as low as 1 1lg/plate (de Lorenzo et al., 1978; Sikka and Florczyk, 19781. Mutagenic activity was dependent upon metabolic activation by hepatic microsomes. Diallate appeared incapable of inducing frameshift mutations. The ultimate mu- tagen produced from diallate is believed to be 2-acrolein (Rosen et al., 1980). Carcinogenicity Diallate is carcinogenic in mice following oral ad- ministration (IARC, 19761. Daily administration of an oral 215-mg/kg bw dose for 4 weeks and then 560 ppm in the diet (approximately 215 ma/ kg bw) for an additional 74 weeks significantly increased the incidence of hepatomas in male and female (C57BL/6 x C3H/Anf)F~ mice and male and female (C57BL/6 x AKR)F~ (strain Y) mice (BRL, 1968a; Innes et al., 19691. The incidence of hepatomas is shown in Table 9-7. Administration of diallate by the subcutaneous route (1 g/kg) for 74 weeks resulted in an increased incidence of systemic reticulum cell sarcomas only in male (C57BL/6 x C3H/Anf)F~ mice (BRL, 1968a). Diallate was found to be noncarcinogenic when administered to rats in the diet at 150 and 300 ppm (15 and 30 mg/kg bw) for 78 weeks (Weisburger et al., 19811.

3 ~ 2 DRINKING WATER AND HEATH Developmental Elects Data are not available. CONCLUSIONS AND RECOMMENDATIONS Diallate is a mutagenic and carcinogenic herbicide. No risk estimate has been calculated, since the data on carcinogenicity of diallate are limited and the early study by Innes et al. (1969) involved a single-dose level using a limited protocol. Diallate should be reevaluated when further carcinogenicity data become available. SU LFALLATE 2-ChIoro-2-propeny' diethy~carbamodithioate CAS No. 95-06-7 RTECS No. EZ5075000 S Cl C2 He \ N—C—S—CH2C = CH2 C2 Hs Sulfallate is a carbamate herbicide used for preemergent control of the growth of grasses and weeds in fruit and vegetable crops. Its molecular weight is 223.79. It is an amber liquid at 25°C, boils at 128-130°C, and has a density of 1.088 g/ml at 25°C. Sulfallate is soluble in most organic solvents and in water up to 100 ppm at 25°C. METABOLISM Data are not available. HEALTH ASPECTS Observations in Humans Data are not available. Observations in Other Species Acute Effects The oral acute LDso for Sulfallate in rats is 850 mg/kg bw (Gosselin et al., 1984~. No acute effects were described. Mutagenicity Sulfallate produces base-pair substitutions in TA100 and TA1535 strains of Salmonella typhimurium et closes as low as 10 ~g/plate

Toxicity of Selected Contaminants 313 in the presence of metabolic activation (de Lorenzo et al., 1978; Sikka and Florczyk, 19781. In S. typhimurium strains TA98 and TA1538, sul- fallate is inactive, suggesting that it is incapable of inducing frameshift mutations even in the presence of metabolic activation (de Lorenzo et al., 1978; Sikka and Florczyk, 19781. The ultimate mutagen produced from sulfallate is believed to be 2-chloroacrolein (Rosen et al., 1980~. Chronic Effects Chronic feeding studies have demonstrated that rats fed 250-ppm concentrations of sulfallate in feed for 6 months develop eye irritation, tubular nephropathy, and hyperkeratosis of the forestomach (NCI, 1978d). Carcinogenicity The carcinogenic potential of sulfallate has been stud- ied in Osborne-Mendel rats and B6C3F~ mice (NCI, 1978d). The com- pound was administered in the diet for 78 weeks. The high and low dietary concentrations (overweighted averages) of sulfallate for each species were, respectively, 410 and 250 ppm for male rats, 404 and 250 ppm for female rats, 1,897 and 949 ppm for male mice, and 11,815 and 908 ppm for female mice. Rats were observed for 25 to 26 weeks after the last dose of sulfallate, whereas mice were observed for 12 to 13 weeks afterward. Sulfallate-treated rats had statistically significant dose-dependent in- creases in incidence of mammary adenocarcinomas (females) and stomach neoplasms (males). In male mice there was an increased incidence of combined alveolar/bronchiolar carcinomas and adenomas, while the in- cidence of mammary adenocarcinomas was elevated in female mice. Fe- male rats appeared to be the most sensitive species, so the mammary tumor data were used to estimate lifetime risk and an upper 95% confidence limit for human exposure. Carcinogenic Risk Estimate Table 9-8 shows the tumor incidence rates used to calculate the risk estimate by the generalized multistage model (Table 9-9), which is described earlier in this chapter. TABLE 9-8 Tumor Incidence in Sulfallate-Exposed Ratsa Animal Tumor Site Dose, ppm (mg/kg/day)b o 250 (11) 404 (17.8) Tumor Rates Osborne-Mendel rat Female Mammary glandc 0/50 7/50 1 1/48 aBased on data from NCI, 1978d. bConverted from ppm in diet. CAdenocarcinoma.

3 ~ 4 OR ~ N K'NG WATER AN D H EALTH TABLE 9-9 Carcinogenic Risk for Sulfallate Estimated with the Generalized Multistage Modela Upper 95% Confidence Estimated Human Estimate of Lifetime Animal Sex Lifetime Riskb Cancer Riskb Osbo~ne-Mendel rat Female 1.0 x 10-6 1.6 x 10-6 aBased on data from NCI, 1978d. bAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ~g/liter. Developmental Effects Data are not available. CONCLUSIONS Sulfallate is a mutagenic and carcinogenic herbicide. It appears to be a more potent carcinogen than diallate, discussed above. Sulfallate pro- duced a broad range of tumors at multiple sites in several organs, whereas diallate produced a limited number of tumors at a single site. However, the carcinogenicity study of diallate was performed using a limited pro- tocol. Dl BROMOCH LOROPROPAN E 1, 2-Dibromo-3-chloropropane CAS No. 96-12-8 RTECS No. TX 8750000 H H H 1 H—C—C—C—H 1 1 Br Br Cl 1,2-Dibromo-3-chloropropane (DBCP) has a molecular weight of 236.4. It is an amber to dark-brown liquid with a boiling point of 196°C and a

Toxicity of Selected Contaminants 3 ~ 5 melting point of 6°C. It is a short-chain aliphatic halogenated hydrocarbon. Technical DBCP, used as a soil fumigant and nematocide since the 1950s, typically contains several impurities such as epichlorohydrin and allyl chloride. Several reviews of the toxicity of DBCP have been published (Babich et al., 1981; IARC, 1979; NIOSH, 1978a; Whorton and Foliart, 1983~. DBCP was evaluated in Volume 4 of Drinking Water and Health (NRC, 1982b, pp. 209-2141. The following section is an examination of data that have become available to the committee since then. HEALTH ASPECTS Observations in Humans Observations regarding the health effects of DBCP in humans have been focused on workers employed in the manufacture and agricultural appli- cation of this nematocidal fumigant (Whorton and Foliart, 19831. Although produced and in use since the 1950s, the chemical's effect on human testicular function was not recognized until 1977. At that time, a study of male DBCP production workers at a California production plant con- firmed workers' concern about apparent infertility by demonstrating azoo- spermia or severe oligospermia in 14 of the 25 men studied (Whorton et al., 1 9771. These findings were later reconfirmed by extended observations at the same plant (Whorton et al., 1979) and in studies of DBCP production workers elsewhere (Egnatz et al., 1980; Lipschultz et al., 1980; Marquez Mayaudon, 1978; Milby and Whorton, 1980; Potashnik et al., 19791. Similar sperm-count reductions were observed in workers exposed during agricultural application of DBCP (Glass et al., 1979; Sandifer et al., 1979~. The degree to which spermatogenesis was reduced appears to have been directly related to the intensity and duration of chemical exposure. Reduction in sperm production is accompanied endocrinologically by elevations in serum levels of follicle-stimulating and luteinizing hormones, and histologically by reduction or absence of spermatogenic cells in tes- ticular seminiferous tubules, while Leydig's and Sertoli's cells remain normal (Biava et al., 1978; Potashnik et al., 19791. Egnatz et al. (1980) concluded that within a population of 232 workers who may have been exposed to DBCP, follicle-stimulating and luteinizing hormone levels were above normal, sperm counts were significantly lower, and testicular vol- ume was less than that of a control population of 97 unexposed workers. From their findings of a significant correlation between the magnitude of DBCP-induced effects and the time since termination of DBCP exposure, the authors concluded that DBCP-induced effects may be revers~ible. The possibility that reduction of spermatogenesis may also involve disturbance of genetic material in sperm was suggested by Kapp et al. (1979), who

3 ~ 6 OR ~ N K! NG WATER AN D H MATH observed an increased frequency of Y-chromosome nondisjunction in the sperm of 18 DBCP-exposed workers (3.8% compared with 1.2% in 15 controls). An interview study of 62 agricultural workers in Israel suggests that spontaneous abortions may have been more frequent in wives of DBCP-exposed workers (19% of 121 pregnancies) than in wives of unex- posed workers (6.6% of 76 pregnancies) (Kharrazi et al., 19801. Follow-up studies of DBCP-exposed workers have demonstrated that recovery of sperm production either does not occur or is greatly impeded in men with a complete absence of spermatogenesis as a result of exposure. For men in whom spermatogenesis is only partially inhibited, complete recovery appears to take place within a few months after cessation of exposure (Lantz et al., 1981; Whorton and Milby, 1980; Whorton et al., 19791. Retrospective analysis of workers' reproductive histories suggests that if history-taking had been more thorough, the presence of impaired fertility in DBCP-exposed workers may have been detected as early as 1959, approximately 18 years before the effect was recognized (Levine etal.,1981,19831. The National Institute for Occupational Safety and Health (NIOSH, 1978a) described pineapple workers monitored for exposure to OBCP (0 to 1.8 ppm during the workday). Sperm counts were not found to vary among groups exposed to different doses in this range. Thus, NIOSH concluded that a 1-ppm exposure has no observable effects on male fer- tility. Whorton and Foliart (1983) reviewed these epidemiological data and concluded that occupational exposure to DBCP at levels significantly above l ppm can lead to sperm and testicular abnormalities. However, they emphasized that the reversibility of those effects and their dose re- sponse have not been well defined. No observations concerning possible toxic effects in women are re- corded in the medical literature. Likewise, no studies have been published regarding possible long-term sequelae such as cancer in humans. Observations in Other Species Acute E.ffects The acute lethality of ingested DBCP has been inves- tigated in a number of species of laboratory animals. Oral LDso values determined by Torkelson et al. (1961) ranged from 180 mg/kg bw in male rabbits to 410 mg/kg bw in female mice. The oral LDso of 300 mg/kg bw in male rats was similar to an LDso of 350 mg/kg bw for rats reported by Rakhamatullaev (1971), who also reported 316 mg/kg bw for guinea pigs and 440 mg/kg bw for rabbits. The dosed animals showed transient ex- citation followed by motor incoordination and impaired sensitivity to pain.

Toxicity of Selected Contaminants 317 The kidney and testes are the primary target organs of DBCP following a single dose of the chemical. Kluwe (1981a) conducted a dose-response study in which male rats were given a single subcutaneous injection of 0, 10, 20, 40, 80, or 120 mg/kg bw and sacrificed 24 hours later. Although no adverse effects were seen in response to the 10- and 20-mg/kg doses, 40 mg/kg significantly altered several urinary indices of kidney injury and caused marked cytoplasmic vacuolization of the renal proximal tubular epithelium in the outer medulla. Forty mg/kg produced only mild hepa- tocellular swelling in the periportal region of liver lobules. The highest dose (120 mg/kg) was required to produce hepatic centrilobular necrosis. Examination of the testes and epididymis revealed relatively minor cellular injury 24 hours after the two highest doses (80 and 120 mg/kg). In a subsequent study of the development and repair of DBCP-induced lesions, Kluwe (1981b) saw progressive seminiferous tubular desqua- mation and atrophy of the testes of rats given a single 100-mg/kg sub- cutaneous injection. Although there was some evidence of recovery, many seminiferous tubules were still devoid of germinal cells, and epididymal sperm density remained low 30 days after dosing. The kidneys contained large areas of fibrotic tissue and foci of interstitial nephritis at 30 days. Reznik and Sprinchan (1975) similarly observed prolonged effects in rats given a single oral dose of 100 mg/kg bw. Pronounced reductions in sperm count and sperm motility, as well as morphological changes of the tes- ticular germinal epithelium, were noted 1 and 2 weeks after exposure. Many studies in which DBCP was given in a single, large oral dose were conducted to investigate biochemical changes and mechanisms of toxicity. Kato et al. (1980b) found a dose-dependent reduction in levels of glutathione (known as GSH) in the livers of male rats given as little as 20 mg/kg bw orally. A sufficient dose of DBCP would be expected to reduce GSH levels, since GSH plays an important role in the detoxification of DBCP. Maximum covalent binding of ~4C-labeled DBCP to liver and kidney macromolecules in rats 6 hours after oral administration of a 20- mg/kg dose coincided with significant depletion of hepatic GSH (Kato et al., 1980a). This dose also resulted in incorporation of radioactivity into liver protein, RNA, and DNA 24 hours after dosing. The findings of Kato et al. (1980a,b) support the concept that GSH protects vital nucleophilic sites in cellular macromolecules of target organs from electrophilic attack by reactive DBCP metabolites. Kluwe et al. (1981), however, discounted the protective role of GSH in DBCP toxicity. These investigators noted that the testes are a major target organ of DBCP, yet the chemical had little effect on testicular nonprotein sulfhydryl (NPS, largely GSH) in mice. They also Qbserved that DBCP caused relatively minor liver injury, yet it markedly depleted hepatic GSH. Kluwe et al. (1981) pointed out that DBCP-induced GSH

3~8 DRINKING WATER AND HEALTH depletion appears to correlate well with organ distribution of the chemical and with the organs' capacities for GSH conjugation. In a subsequent study in Fischer 344 rats, Kluwe et al. (1982) found that subcutaneous injection of DBCP did not lower GSH levels in the kidneys or testes. The investigators did find that pretreatment with di- ethylmaleate lowered hepatic, renal, and caput epididymal NPS. Single subcutaneous injections of DBCP produced dose-dependent lesions in the liver, kidney, testes, and caput epididymis. Diethylmaleate treatment 90 minutes before DBCP treatment enhanced the nephrotoxic potency of DBCP. Seminiferous tubular degeneration, as determined 48 hours after DBCP treatment, was greater in rats pretreated with 600-mg/kg doses of diethylmaleate than in nonpretreated controls. These results indicate that DBCP is a depletor of hepatic and caput epididymal NPS in the acutely toxic dose range. Since NPS concentrations were not lowered in two of the major target organs, the kidney and testes, acute DBCP injury would not appear to be dependent on local GSH depletion. However, the greater susceptibility of kidney and testes to DBCP injury after pretreatment with diethylmaleate suggests an important role for NPS in modulating DBCP toxicity, especially in the liver. Single, large oral doses of DBCP produce a number of effects in a variety of tissues. Moody et al. (1982a) found a significant reduction in the number of microsomes in the cytochrome P450 isolated from liver, kidney, testes, lung, and small intestine of male Sprague-Dawley rats 48 hours after the animals received 200 mg/kg orally. There was also a slight decrease in hepatic microsomal NADPH cytochrome-c reductase activity and cytochrome-b5 content, but no apparent change in dealkylation activ- ities. Moody et al. (1982b) evaluated the influence of a 200-mg/kg bw oral dose on liver protein synthesis in male Sprague-Dawley rats. No effect was seen on total liver DNA content, on RNA, or on protein synthesis. Although DBCP did not cause a change in overall protein synthesis, the authors noted that their findings did not rule out the action of DBCP on the synthesis of specific proteins in the liver. Suzuki and Lee (1981) gave male Sprague-Dawley rats a 125-mg/kg oral dose of DBCP and then measured aryl hydrocarbon hydroxylase (AHH) and epoxide hydrolase activities in selected tissues. AHH was not induced in the liver or prostate, but it was induced in the stomach, testes, and kidneys. Epoxide hydrolase was maximally increased 72 hours after dosing in the liver, kidney, testes, and prostate. Moody et al. (1981) assessed the influence of a 600-mg/kg bw oral dose on a series of hepatic microsomal parameters in male Sprague-Dawley rats in order to gain insight into the mechanism of DBCP-induced decrease in cytochrome P450. This very high dose did not alter the content of hepatic microsomal proteins, RNA, phospholipids, diene conjugates, NADPH cytochrome-c

Toxicity of Selected Contaminants 319 reductase activity, or cytochrome-b5, but it did significantly diminish cytochrome P450 levels. The decrease in cytochrome P450 was accom- panied by changes in microsomal fatty acid composition, leading the authors to speculate that the decrease may have resulted from DBCP- induced perturbation of the lipid environment of cytochrome P450 in the microsomes. Tofilon et al. (1980) offered another explanation for the drop in cyto- chrome P450. They observed a maximum decrease in testicular micro- somal cytochrome P450 in male Sprague-Dawley rats 2 days after oral administration of 200 mg/kg bw and a return to normal cytochrome P450 levels after 6 days. Heme oxygenate and b-aminolevulinic acid synthetase activities were not altered 24 hours after dosage, but heme synthesis was significantly inhibited. These results indicate that DBCP reduces cyto- chrome P450 levels by inhibiting heme synthesis through action at a site other than 8-aminolevulinic acid synthetase in the heme biosynthesis path- way. The use of microsomal enzyme inducers and inhibitors to elucidate the role of metabolism in DBCP toxicity has yielded conflicting results. Kato et al. (1980b) found that phenobarbital pretreatment of rats exposed to i4C-labeled DBCP resulted in increased elimination of radioactivity in urine, enhanced GSH depletion in liver and kidney, and increased binding of the radiolabeled compound in liver and kidney macromolecules, as well as greater liver damage and increased mortality. Pretreatment with SKF- 525A had the opposite effect on each parameter. Kluwe et al. (1981) reported that pretreatment of male ICR mice with enzyme inducers- polybrominated biphenyls (PBBs) or with the enzyme inhibitor piper- onyl butoxide decreased the GSH-depleting action of DBCP in the liver and kidneys. SKF-525A was without demonstrable effect. Pretreatment with PBBs resulted in an increase in the LDso of DBCP in mice, but piperonyl butoxide was without effect. In a subsequent paper, Kluwe (1983) reported that phenobarbital pretreatment reduced DBCP nephro- toxicity, hepatotoxicity, seminiferous tubular atrophy, and degeneration of the epithelium of the caput epididymis in rats. The influence of other microsomal enzyme inducers and inhibitors on DBCP toxicity varied from one organ to another, leading the author to point out the complexity of interpreting the effects of these agents on extrahepatic toxicity. In another study of the role of metabolism in DBCP toxicity, Kluwe et al. (1983) examined the relative toxicity of DBCP, epichlorohydrin, a- chlorohydrin, and oxalic acid to the liver, kidney, testes, and epididymis. DBCP, epichlorohydrin, and (x-chlorohydrin caused similar lesions in the epididymis and testes of rats. The morphological and functional effects of the four compounds on the kidney differed, suggesting that DBCP nephropathy does not result from the metabolism of DBCP to epichlo-

3 2 0 OR ~ N K! NG WATER AN D H EALTH rohydrin or cx-chlorohydrin or from the subsequent metabolism of these two compounds to oxalic acid. Subacute Elects A number of studies on subacute effects have been conducted with oral doses of DBCP. Kluwe (1981a) gave 40 mg/kg of analytical-grade (~99% pure) DBCP daily for 4 days to male rats by Savage. This dosage regimen produced significant functional and histo- pathological changes of the kidneys as well as marked testicular and epididymal degeneration. Relatively little effect was apparent in the liver. Technical-grade DBCP was given by gavage 5 times weekly for 6 weeks to male and female mice and rats in order to establish the maximally tolerated doses to be used in a subsequent carcinogenicity bioassay (NCI, 1978a). The mice and rats received doses of 100 to 631 and 25 to 160 mg/kg/day, respectively. Reduced body weight gain was seen in male and female rats at the lowest dosage level (25 mg/kg). Deaths occurred in groups of female rats receiving doses greater than 25 mg/kg as well as in male rats receiving more than 40 mg/kg. The mice were less sensitive to DBCP, in that a dose of 398 mg/kg was required to cause death of males and females. Several investigators have conducted long-term oral dosing studies of DBCP. Torkelson et al. (1961) maintained male and female rats on pow- dered diets containing DBCP concentrations of 0, 5, 20, 50, 150, 450, or 1,350 ppm for 90 days. Decreased body weight gain was manifest in the females receiving 150 ppm and in animals of both sexes receiving 450 and 1,350 ppm. Female rats ingesting '20 ppm exhibited increased kidney-to-body-weight ratios. If the rats consumed a quantity of food equal to 10% of their body weight daily, the animals receiving 20 ppm would ingest 1 mg/kg bw daily. The lowest-observed-effect level (LOEL) in this study thus was 1 mg/kg bw, and the no-observed-effect level (NOEL) was 0.25 mg/kg bw. In another study, Reznik and Sprinchan (1975) gave male and female rats 10 mg/kg bw orally for 4 to 5 months. Decreases in sperm count and sperm motility were initially seen during the second month. These effects became more pronounced as the study progressed. Inhibition of spermatogenesis was observed upon examination of the testes. Progressive alterations of a number of biochemical parameters in testicular and ovarian tissue were reported as were increasingly severe disturbances of the menstrual cycle of the female rats. Subacute inhalation studies have demonstrated that the kidneys, testes, and respiratory tract are particularly susceptible to injury from inhaled DBCP. Torkelson et al. (1961) exposed male and female rats, guinea pigs, and rabbits 7 hours daily to vapor containing 12-ppm concentrations of DBCP 50 to 66 times within 70 to 92 days. The male and female rats experienced 40% to 50% mortality. Cloudy swelling of the proximal

Toxicity of Selected Contaminants 32 ~ tubular epithelium and a slight increase in interstitial tissue were observed in the kidneys of male rats. The most striking changes seen at autopsy were severe atrophy and degeneration of the testes of all three species. In a follow-up experiment reported at the same time, male rats were exposed 7 hours daily to 5-ppm concentrations of DBCP 50 times in 70 days. Histopathological changes were focal: they were limited to the testicular epithelium, the collecting tubules of the kidneys, and the bronchioles. Saegusa et al. (1982) also found that the testicles, kidneys, and lungs were severely injured in male rats subjected continuously for 14 days to vapor containing a 10-ppm DBCP concentration. Testes completely atro- phied with irreversible aspermatogenesis. Reznik et al. (1980b) conducted a more definitive dose-response study in which emphasis was placed on respiratory pathology. Male and female rats and mice were exposed to 0-, 1-, 5-, or 25-ppm concentrations of DBCP by inhalation 6 hours/day, S days/week for 13 weeks. The severity and incidence of histopathological changes of the nasal cavity were dose related. Changes in all dosage groups in the region of the respiratory turbinates included cytomegaly of the basal cells, focal hyperplasia, squamous metaplasia and disorientation of basal and ciliated cells, and loss of cilia. Necrosis and squamous metaplasia of the olfactory, tracheal, and bronchial epithelium were present in the an- imals receiving 25 ppm. Chronic Elects Rakhamatullaev (1971) conducted a study in which male rats were given DBCP at 0, 0.005, 0.05, 0.5, and 5.0 mg/kg bw orally for 8 months. The doses of 0.5 and 5.0 mg/kg were reported to be gonadotoxic, to produce functional disturbances of the liver and kidneys, to alter the composition of the blood, and to cause disturbances in con- ditioned reflexes. The author concluded that 0.005 mg/kg was a NOEL and that 0.05 mg/kg was a LOEL. The latter dosage level was reported to impair conditioned reflexes and reduce the capacity of neutrophils from heat-stressed animals to digest bacteria. No data were presented for most parameters, so it is not possible to assess the magnitude, the statistical significance, or the toxicological significance of the effects reported by Rakhamatullaev. The chronic oral toxicity of DBCP was assessed in conjunction with a National Cancer Institute (NCI, 1978a) carcinogenicity bioassay. Male and female Osborne-Mendel rats and B6C3F~ mice were dosed orally with technical-grade DBCP (purity >90%) 5 times weekly for up to 78 weeks. It was necessary to terminate the study early due to excessive death rates. The time-weighted average doses were as follows: 0, 15, and 29 mg/kg bw for male and female rats; 0, 114, and 219 mg/kg bw for male mice; 0, 110, and 209 mg/kg bw for female mice. There were dose-related decreases in body weight gain and survival of male and female rats. Toxic

322 DRINKING WATER AND H"LTH nephropathy was seen in virtually 100% of the male and female rats at both dosage levels, but not in controls. Microscopically, the nephropathy was characterized by cloudy swelling, fatty degeneration, and necrosis of the proximal convoluted tubular epithelium at the junction of the cortex and medulla. The male and female mice experienced a dose-dependent Increase In mortality. The incidence of toxic nephropathy in the mice was also dose dependent. . . Mutagenicity Bites et al. (1978) investigated the mutagenicity of DBCP by using the Ames reverse mutation assay. They concluded that in the absence of S9 activation in rats pretreated with Aroclor, the mutagenic capability of standard DBCP preparations (0 to 1,600 ~g/plate) was due solely to epichlorohydrin, which was included as a stabilizer. However, after the addition of S9, technical-grade and highly purified DBCP (20 to 200 ungulate) were equally mutagenic. On the basis of those data, Biles et al. (1978) concluded that DBCP is a potent indirect mutagen in bacteria. That conclusion has been confirmed by recent work of Moriya et al. (1983), who found DBCP to be mutagenic to Salmonella typhimurium TA1535 and TA100 and to Escherichia cold WP2 her. Zimmering (1983) fed 0.2-mg/ml concentrations of DBCP in 0.01% ethanol or ethanol alone to Canton-S male Drosophila melanogaster for 72 hours and then to individual males mated with Base females. DBCP treatment produced sex-linked recessive mutations in 9.5% of the first brood. In Drosophila, it also caused loss of X or Y chromosomes and induced increases in heritable translocations. DBCP has been shown to induce sister-chromatic exchange (SCE) and chromosome aberrations in cultured Chinese hamster ovary cells over a range of applied doses (Tezuka et al., 1980~. The authors concluded that, compared with 10 other compounds in the same study, DBCP-induced SCE and chromosome aberrations correlate well with DBCP's relative potency in a bacterial assay for mutation. Kapp et al. (1979) found that semen samples of 18 workers exposed to DBCP had higher numbers of sperms containing two Y-chromosomes than did nonexposed subjects. Based on an analysis of the aneuploid karyotypes detected, the authors suggested that environmental exposure to DBCP can produce irreversible genetic change in humans. The mechanism by which DBCP is metabolically activated to a toxicant in mammalian cells has not been well characterized. Recently, however, Kluwe (1983) has shown that primary tissue damage in rats is enhanced by previous exposure to 3-methylcholanthrene, which is an inducer of AHH activity. Therefore, the author suggested, DBCP may be processed by the cytochrome's P450 enzyme pathway. Additional work is required to establish such a mechanism unambiguously.

Toxicity of Selected Contaminants 323 Biochemical data suggest that DNA is a principal target for DBCP in mammalian cells. Lee and Suzuki (1979) showed that a single intraperi- toneal 100-mg/kg injection of DBCP to prepubertal male mice induced significant unscheduled DNA synthesis (DNA repair) in premeiotic germ cells but not in spermatozoa. The authors concluded that in premeiotic germ cells (but not adult spermatozoa), metabolites of DBCP may attack DNA and then induce an excision-DNA repair mechanism. Work is re- quired to verify those initial conclusions, but the data suggest that in mammalian cells, DBCP may be activated in a manner similar to other hydrocarbon carcinogens. Carcinogenicity In the NCI (1978a) bioassay, technical-grade DBCP (90% pure) containing 16 impurities was given by gavage in corn oil to male and female Osborne-Mendel rats and B6C3F~ mice. Time-weighted average dosages for male and female rats were 0, 15, and 29 mg/kg bw. The time-weighted averages were 0, 114, and 219 mg/kg for male mice and 0, 110, and 209 mg/kg for female mice. In this bioassay, technical- grade DBCP induced a high incidence of squamous cell carcinomas of the forestomach and toxic nephropathy in male and female rats and mice and a high incidence of mammary adenocarcinomas in female rats. Van Duuren et al. (1979) found that DBCP behaved as a strong initiator of carcinogenesis in an epidermal initiation-promotion assay with 12-O- tetradecanoylphorbol-13-acetate (TPA) in mice. Applied to mouse skin alone, DBCP did not produce skin tumors, but it increased substantially the incidence of lung and stomach tumors. The authors concluded that DBCP is an efficient initiator of epidermal papillomas, but it is not a complete carcinogen in the skin (although it appears to be a complete carcinogen with respect to benign papillomas of the lung and squamous cell carcinomas of the stomach). Reznik et al. (1980a) found that when mice were exposed to 0, 0.6, or 3.0 ppm DBCP for 6 hours/day, 5 days/week for 103 weeks by inhalation, they developed alveolar-bronchial adenomas and carcinomas. Females appeared to be more sensitive, but for both sexes, tumors appeared above background at doses as low as 0.6 ppm. Evidently, DBCP is a strong, complete carcinogen for the noses and lungs of mice. Tumors that could be attributed to DBCP exposure were not detected at distant sites. The National Toxicology Program (NTP, 1982a) conducted an inhalation car- cinogenicity bioassay of technical-grade DBCP containing trace amounts of epichlorohydrin and 1,2-dibromoethane in male and female Fischer 344 rats and B6C3F~ mice. The animals were exposed by inhalation to 0.6- or 3.0-ppm concentrations of DBCP 6 hours/day, 5 days/week for 76 to 104 weeks. DBCP induced significantly higher incidences of tumors of the nasal cavity and tongue in male and female rats and cortical adenomas

324 DRINKING WATER AND HEALTH TABLE 9-10 Tumor Incidence in Rats Gavaged with Dibromochloropropanea Animal Tumor Sex Site Dose (mg/kg/day) Osborne-Mendel rat Male Forestomach 0 15 29 Tumor Rates 0/39 47/50 47/50 aBased on data from NCI, 1978a. in the adrenal glands of female rats. In male and female mice, it induced significantly higher incidences of nasal cavity and lung tumors. Carcinogenic Risk Estimate Results of the NCI (1978a) study, in which DBCP was found to elicit squamous cell carcinomas of the fore- stomach in rats, were used to estimate carcinogenic risk. The tumor in- cidence rates in Table 9-10 were used to make statistical estimates of both the lifetime risk and an upper 95% confidence limit for the lifetime nsk. The risk estimates are expressed as a probability of cancer after a lifetime daily consumption of 1 liter of water containing the compound at a con- centration of 1 ~g/liter. By using the data in the NCI (1978a) study, the committee calculated the lifetime risk and upper 95% confidence bounds shown in Table 9-11 for males who daily consume 1 liter of drinking water containing 1 fig of DBCP per liter. Developmental Elects Ruddick and Newsome (1979) dosed pregnant rats with 0, 12.5, 25, or 50 mg/kg bw orally on days 6 through 15 of gestation. DBCP was not teratogenic, but it produced a dose-dependent reduction in maternal body weight gain. Reproductive Effects DBCP appears to alter male fertility in a signif- icant way. Saegusa et al. (1982) found that the testicles were severely TABLE 9-11 Carcinogenic Risk for Dibromochloropropanea Estimated with Generalized Multistage Model Animal Estimated Human Sex Lifetime Riskb 7.8 x 10-6 Upper 95% Confidence Estimate of Lifetime Cancer Riskb Osborne-Mendel rat Male 9.9 x 10-6 aBased on data from NCI, 1978a. bAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ~g/liter.

Toxicity of Selected Contaminants 325 injured in a study of male rats subjected continuously for 14 days to vapor containing DBCP at 10 ppm. Testicles completely atrophied with irre- versible aspermatogenesis. In a second experiment, Saegusa et al. (1982) exposed male rats continuously for 14 days to 0-, 0.3-, 1-, 3-, or 8-ppm concentrations of DBCP vapor. No testicular lesions were observed at 1 or 15 days after exposure in the 0.3-ppm or 1-ppm exposure groups. A slight decrease in germ cells and atrophy of a few seminiferous tubules were observed in the testes of one of two rats examined in the 3-ppm group. Severe testicular lesions were present in the rats receiving 8 ppm. Rao et al. (1982) conducted a comprehensive inhalation dose-response study of gonadotoxic effects of DBCP. Male rabbits were exposed to DBCP in vapor at 0, 0.1, 1.0, or 10 ppm 6 hours/day, 5 days/week for up to 14 weeks. The time of onset and severity of decreases in sperm count and sperm viability were dose dependent. Under the conditions of this study, 0.1 ppm can be considered the NOEL and 1.0 ppm the LOEL, although the 0.1-ppm animals had an equivocal increase in abnormal sperm count at 14 weeks. The adverse effects on spermatogenesis were largely reversible in the rabbits exposed to DBCP at 1 ppm. Partial recovery from the severe effects of the 10-ppm exposure occurred by the end of a 38-week recovery period. Kluwe et al. (1983) showed that a single subcutaneous treatment of Fischer 344 rats with a 100-mg/kg dose of DBCP decreased male fertility 2 to 7 days after exposure without affecting mating frequency. Infertility was found to be accompanied by a decrease in the rate of glucose oxidation in mature sperm. The authors concluded that among its other possible effects, exposure to DBCP may alter the overall metabolism of post- testicular sperm in rats. Kluwe (1981b) has also presented data suggesting that among other tissue damage, single 100-mg/kg subcutaneous doses of DBCP induce generalized testicular damage: seminiferous tubular desquamation and atrophy and substantial cell death occurred after subcutaneous treatment was initiated in rats. On the other hand, Osterloh et al. (1983) exposed mice to increasing doses (5 to 300 mg/kg/day) of DBCP subcutaneously and found no evidence of DBCP-induced changes in sperm count, mor- phology, or overall testes weight. The authors concluded that DBCP- induced effects may be species specific. CONCLUSIONS AND RECOMMENDATIONS DBCP is mutagenic and carcinogenic to rats and mice in several organ systems. It also has adverse effects on human male fertility. As there are no adequate feeding studies showing adverse reproductive effects with clear dose-response relationships, a risk assessment for repro- ductive effects would be inappropriate. Although good inhalation studies

326 DRINKING WATER AND H"LTH have been conducted, the committee decided not to use data from them to extrapolate to the oral route for this compound. Nevertheless, DBCP would be a good substance to use with the extrapolation model developed in Chapter 7 when additional phannacokinetic values are experimentally determined. CHLOROPROPANES AND CHLOROPROPENES 1,2-DICH LOROPROPAN E CAS No. 78-87-5 RTECS No. TX9625000 1,2,3-TRICHLOROPROPANE CASNo.96-18-4 RTECS No. TZ9275000 H H 1 1 CH3—C—C—H 1 1 C1 C1 H H H 1 1 1 H—C—C—C—H 1 1 1 C1 C1 C1 1,3-DICHLOROPROPENE (cis and bans) CAS No. 542-75-6 RTECS No. UC8310000 C1 H' C1 \C=C AH ~ CH2 C1 - CH2 C1 Ez H' —H The chloropropanes and chloropropenes (CPs) are generally found in mixtures used as soil fumigants and fungicides. Following the discovery

Toxicity of Selected Contaminants 327 of the testicular toxicity and carcinogenicity of dibromochloropropane (DBCP) and the acute toxicity and carcinogenicity of ethylene dibromide (EDB) (see discussion later in this chapter), the toxicological literature for the CPs has generally concerned combinations of two or three of these compounds, or one of the compounds in combination with a brominated analog. Two combinations of 1,2-dichloropropane (1,2-DCP) (30%), cis- 1,3-dichloropropene (28%), and trans-1,3-dichloropropene (27%) are commercially available as D-D (a broad-spectrum soil fumigant that con- sists primarily of cis-1 ,3-dichloropropene, trans-1 ,3-dichloropropene, and 1,2-pop) and CBP-55 (primarily 1-chloro-3-bromo-1-propene and 1,2- dichloropropane t40%~) (Hine et al., 1953; Hutson et al., 19711. METABOLISM Following oral administration of 1,2-dichlorof1-~4C]propane orcis- or trans-1,3-dichlorof2-~4C]propene to rats, between 80% and 90% of the radioactivity was eliminated in the feces, urine, and expired air within 24 hours (Hutson et al., 19711. When rats were exposed to varying concen- trations (30 to 900 ppm) of a vapor containing 91% 1,3-dichloropropene (both isomers), absorption was not observed to increase proportionately with increasing exposure level, partly because of a 40% to 50% depression in respiratory minute volume at the higher dose levels. Postexposure elim- ination curves for both isomers displayed an initial rapid and dose- dependent elimination phase (2- to 3-minute half-life), followed by a slower elimination phase (38- to 44-minute half-life) (Stott et al., 19851. The metabolism of 1,2-pop involves the formation of 1-chloro-2- hydroxypropane. This intermediate metabolite is detoxified via epoxida- tion to 1,2-epoxypropane, which can either conjugate with glutathione or be hydrolyzed to propane-], 2-diol. Conjugation would lead to the ultimate excretion of the major urinary metabolite, N-acetyl-S-~2-hydroxy- propyl~cysteine. Its hydrolysis to propane-1,2-diol and subsequent oxi- dation to lactate would lead to its complete oxidation-to carbon dioxide in the tricarboxylic acid cycle. Due to the volatility and lipophilic nature of 1,2-pop, some is excreted via the lungs, but no unchanged compound is excreted in the urine (Jones and Gibson, 19801. The cis isomer of 1,3-dichloropropene has been shown to be metabolized by conjugation with glutathione. Seventy-six percent of an oral dose appeared within 24 hours in the urine of rats as the mercapturic acid N-acetyl-S-(cis-3-chloroprop-2-enyl~cysteine (Climie et al., 19791. This compound is also the major urinary metabolite of 1,3-dichloropropene in mice (Dietz et al., 1984) and has been detected in the urine of humans exposed to cis-1,3-dichloropropene vapor during field applications (Osterloh et al., i9841.

328 DRINKING WATER AND H"LTH HEALTH ASPECTS Observations in Humans Published observations concerning humans exposed to 1,2-pop consist of case reports of acute illness resulting from accidental exposure through inhalation, ingestion, or dermal contact. In Italy, four cases of acute liver damage (one fatal) were observed in persons ingesting unknown quantities of the commercial solvent "trilene," 70% to 100% of which is 1,2-pop (Chiappino and Secchi, 1968; Secchi and Alessio, 1968; Secchi et al., 19681. Contact dermatitis on hands and feet (not described more fully) was found in two women one who worked for 6 years and one for 4 years at two separate plastics production plants in Poland (Grzywa and Rudzki, 1981~. Dermatitis was found in workers using D-D as a soil fumigant (Bodansky, 19371. 1,2-pop has also been reported to cause headache, vertigo, tearing, and irritation of the mucous membranes when used as a dry-cleaning agent. No concentrations were given for inhalation exposure from dry-cleaning or from use as a fumigant (Hine et al., 19531. Venable et al. (1980) examined the fertility of male workers engaged in the manufacture of glycerine with potential exposure to allyl chloride, epichlorohydrin, and 1,3-dichloropropene. Sperm counts and normal sperm forms were determined for each worker and then compared with controls. The study included 79 volunteers out of a possible 123 workers, giving a volunteer rate of 64%. The authors reported that there was no association between decreased fertility in workers and workplace exposure to any of the three substances. No other studies assessing potential delayed or chronic health effects associated with human exposure to 1,2-pop are recorded in the literature. There are no published studies on human health effects from exposure to 1,2,3-TCP and 1,3-dichloropropene. Observations in Other Species Acute Effects The acute lethality of 1,2-pop has been studied in a number of animal species. Oral LDso values of 860 mg/kg bw and 2,200 mg/kg bw have been reported, respectively, for mice (Windholz et al., 1976) and rats (Smyth et al., 1969~. S myth et al. (1969) also reported that three of six rats died from inhaling 2,000-ppm vapor concentrations of 1,2-pop for ~ hours. Heppel et al. (1946) observed that 2,200-ppm vapor concentrations of 1,2-pop caused marked irritation of ocular mu- cous membranes of guinea pigs as well as marked inebriation in rats and mice. The authors subjected guinea pigs, rats, mice, rabbits, and dogs to

Toxicity of Selected Contaminants 329 a single 7-hour exposure to each of a series of vapor concentrations of 1,2-pop. Mice were the most sensitive. A1126 mice exposed to a con- centration of 1,000 ppm died within 4 hours after inhalation. The liver, kidneys, and adrenals appear to be the primary target organs for acute injury by 1,2-pop. In an early investigation of the anthelmintic properties of a series of chlorinated alkyl hydrocarbons, Wright and Schaf- fer (1932) observed gross and microscopic lesions in the livers of 11 dogs given a single oral dose of 230 to 580 mg/kg bw. Microscopic changes included parenchymatous degeneration, fatty vacuolation, and cloudy swelling or edema. The 1,2-pop was said to have been obtained from a commercial source and purified by distillation. Drew et al. (1978) exposed male CD1 rats for 4 hours by inhalation to 1,000-ppm concentrations of 1,2-pop. Liver injury was evidenced at 24 and 48 hours after exposure by parallel increases in serum levels of glutamic- oxaloacetic transaminase (SGOT), glutamic-pyravic transaminase (SGPr), and ornithine carbamyl transferase (OCT) activities. The activity of each of the three enzymes was higher at 48 hours than at 24 hours. Highman and Heppel (1946) examined the development and regression of lesions in Sprague-Dawley rats and in guinea pigs exposed by inhalation to 2,200-ppm concentrations of commercial-grade 1,2-pop for 7 hours. The results of fractional distillation studies led the authors to conclude that their 1,2-pop probably contained a considerable percentage of other isomers. The authors found centrilobular necrosis and diffuse to midzonal fatty degeneration in the livers of rats 24 hours after exposure. Groups of rats sacrificed on subsequent days after the single exposure had progressive resolution of the hepatic lesions as well as diminished fatty vacuolation of the renal convoluted tubular epithelium. Fatty vacuolation of the liver and kidneys was also observed in the guinea pigs underdoing 7-hour inhalation exposures up to 2,200 ppm two to five times. ~n acct~t~on, tnese guinea pigs exhibited necrosis of the adrenal cortex affecting all but a narrow outer rim of the tissue. The adrenal cortical cells rapidly regen- erated during a 16-day recovery period. Proliferation of fibroblasts and formation of a thin layer of connective tissue around the adrenal medulla were also noted during this time. Heppel et al. (1946) found histopatho- logical lesions in mice that died after being exposed for 2 hours to 1,500 ppm or for 4 hours to 1,000 ppm 1,2-pop by inhalation. The lesions included hepatic centrilobular necrosis and fatty degeneration of the liver and kidneys. No data pertaining to the acute toxicity of 1,3-dichloropropene were located in the literature. However, Hine et al. (1953) did evaluate the acute toxicity of technical-grade D-D, which is a broad-spectrum soil fumigant that consists primarily of cis-1,3-dichloropropene, trans-1,3- dichloropropene, and 1,2-pop. They reported that the oral LDso for O O . .

330 DRINKING WATER AND HEALTH D-D was 300 mg/kg bw in inbred male Swiss mice and 140 mg/kg bw in Long-Evans rats (sex not reported). The animals exhibited hyperexcit- ability followed by tremors, incoordination, depression, and breathing difficulties. Fatty degeneration of the liver and hemorrhage of the lungs were occasionally found in animals that died several days after oral dosing. The 4-hour inhalation LC50 for D-D in rats was calculated by the authors to be 1,000 ppm. Rats that died from inhaling the vapors exhibited severe pulmonary edema with varying degrees of interstitial and alveolar hem- orrhage. Undiluted D-D was also found to be a severe skin and eye irritant and to cause erosion and occasional hemorrhage of the gastrointestinal mucosa when administered intragastrically. Little information on the acute toxicity of 1,2,3-TCP was found in the scientific literature. In an early paper, Wright and Schaffer (1932) reported that 1,2,3-TCP was the most toxic of a series of 18 chlorinated alkyl hydrocarbons that they tested in dogs. The 1,2,3-TCP was synthesized by the investigators and purified by distillation. Three dogs (sex and strain unspecified) were fasted for 12 hours before intragastric administration of the chemical. One dog received 0.2 ml/kg, one received 0.3 ml/kg, and one received 0.5 ml/kg. All three were completely narcotized within 1 to 2 hours and died within 1 to 2 days of dosing. Microscopic examination of the livers revealed fatty vacuolation and centrilobular necrosis. Cloudy swelling, desquamation of the tubular epithelium, and nuclear changes indicative of beginning necrosis were observed in the kidneys. In a more recent study, Drew et al. (1978) found that spectrograde (99% pure) 1,2,3- TCP produced liver damage in male CD1 rats. Rats that inhaled 500-ppm concentrations of the chemical for 4 hours exhibited increased levels of SOOT, SGPT, and OCT 24 and 48 hours after exposure. Concurrent 4- hour inhalation exposure to 1,2,3-TCP at 500 ppm and 1,2-pop at 1,000 ppm resulted in additive increases in SGOT and OCT. By 48 hours after exposure, the increases were significantly less than additive. Subacute Elects It appears that the only subacute studies in which 1,2-pop has been given orally were conducted as part of an NTP (1983a) carcinogenesis bioassay. In a preliminary range-finding study, Fischer 344/N rats and B6C3F~ mice of each sex were given reagent-grade (99.4% pure) 1,2-pop in corn oil by gavage in doses of 0, 125, 250, 500, 1,000, or 2,000 mg/kg daily for 14 consecutive days. All rats receiving 2,000 mg/kg died. The two highest doses were lethal to all but one of the male and female mice receiving the 2,000- and 1,000-mg/kg doses. Three of five male mice receiving 500 mg/kg died. A follow-up study was conducted to determine the doses of 1,2-pop to be used in a chronic carcinogenicity bioassay. B6C3F~ mice of each sex were given reagent-grade 1,2-pop in corn oil by gavage in doses of

Toxicity of Selected Contaminants 331 0, 30, 60, 125, 250, or 500 mg/kg, 5 days/week for 13 weeks. Male and female Fischer 344/N rats similarly received doses of 0, 60, 125, 250, 500, or 1,000 mg/kg. All male and female rats given 1,000 mg/kg died, as did 5 of 10 males in the 500-mg/kg group. Histopathological exami- nation revealed fatty changes, centrilobular congestion, and centrilobular necrosis in the livers of some of the rats receiving 1,000 mg/kg. There were no apparent increases in mortality or histopathological changes in mice at any dosage level. Heppel et al. (1946) conducted a series of experiments in which guinea pigs, rats, rabbits, and dogs were exposed to commercial-grade 1,2-pop by repeated daily inhalation exposures. The investigators noted that their test materials probably contained a considerable percentage of other iso- mers. The animals were subjected to 1,000-, 1,600-, or 2,200-ppm vapor concentrations of 1,2-pop for 7 hours/day, 5 days/week for as many as 128 exposures. The lowest concentration (1,000 ppm) caused some deaths among all animal species except rabbits during the course of the exposure regimens. Liver, kidney, and adrenal injuries were most pronounced dur- ing the initial days of the study. Animals that did not succumb to the chemical became resistant to its injurious effects. Guinea pigs that survived 26 exposures to 1,000-ppm concentrations developed no significant his- tological changes other than a subcortical layer of fibrous tissue in the adrenals. Rats sacrificed after as many as 97 exposures to 1,000 ppm exhibited few changes other than splenic hemosiderosis. In a follow-up histopathological study, Highman and Heppel (1946) confirmed that surviving rats and guinea pigs become resistant upon re- peated exposure to concentrations of 1,2-pop. Sprague-Dawley rats and guinea pigs inhaled 2,200-ppm concentrations of 1,2-pop 7 hours daily for 5 consecutive days. Daily sacrifices revealed that lesions in the primary target organs (liver, kidneys, and adrenals) in each species were most severe at 3 days. Resolution was seen thereafter despite continuation of the daily inhalation exposures. The results of subsequent studies indicate that inhalation of relatively low concentrations of 1,2-pop can produce detrimental effects in some species. Heppel et al. (1948) exposed male and female rats and guinea pigs and female dogs 7 hours/day, 5 days/week for approximately 6 months to 400 ppm of commercial-grade 1,2-pop. Groups of animals were sac- rif~ced at intervals during the study and examined for histopathological changes No ill effects were attributed to the exposures other than de- creased body weight gain in the male and female rats. The authors also subjected C57-strain mice to daily 7-hour inhalation exposures to 400- ppm concentrations of 1,2-pop. Most of the mice died during the course of the study. Thus, the mouse appears to be particularly sensitive to inhaled 1,2-pop.

332 DRINKING WATER AND H"LTH Sidorenko et al. (1976) exposed male rats by inhalation continuously for 7 days to 217- to 434-ppm concentrations of 1,2-pop. The investi- gators stated that "verifiable" changes in blood catalases and acetylcho- linesterase were first observed after 4 hours in the 434-ppm group and after 24 hours in the 217-ppm group. Because no data were presented, the magnitude and significance of these changes cannot be ascertained. The authors also reported histological and functional changes in the liver and kidneys. Unfortunately, they did not provide detailed data on these changes, nor did they indicate which vapor levelks) produced the injury or note when the changes occurred following exposure. There is very little information on the subacute toxicity of 1,2,3-TCP. Sidorenko et al. (1976) exposed male rats by inhalation continuously for 7 days to 350-mg/m3 (58-ppm) or 800-mg/m3 (133-ppm) concentrations of 1,2,3-TCP. The exposures were reported to produce changes in blood catalases and acetylcholinesterase as well as "histostructural" evidence of liver and kidney injury. Unfortunately, as for 1,2-pop, the researchers did not provide specific data or give an adequate description of their findings. Torkelson and Oyen (1977) investigated the effects resulting from re- peated inhalation of 1,3-dichloropropene. The sample they used in their study contained 46% cis- 1,3-dichloropropene, 53% trans- 1,3-dichloro- propene, and 1% other chemicals, including epichlorohydrin. In a prelim- inary experiment, repeated inhalation exposure of male and female rats and guinea pigs to 11 or 50 ppm produced gross evidence of liver and kidney damage. Male rats were then exposed to 3 ppm for 0.5, 1, 2, or 4 hours/day, 5 days/week for 6 months. The only adverse effect seen was very slight cloudy swelling of the renal tubular epithelium in the 4-hour exposure group. In a subsequent experiment, 24 male and 24 female rats, 12 male and 12 female guinea pigs, 3 male and 3 female rabbits, and 2 dogs were exposed 125 to 130 times over 185 days to 1 or 3 ppm. Male rats exposed to 3 ppm exibited cloudy swelling of the renal tubular epi- thelium, whereas female rats exposed to 3 ppm had a slight but statistically significant increase in liver-to-body-weight ratio. No changes were ob- served in animals inhaling 1 ppm. On the basis of the study results of Torkelson and Oyen (1977), 1 ppm appears to be a NOEL, whereas 3 ppm appears to be a LOEL in rats exposed to 1,3-dichloropropene by inhalation. Parker et al. (1982) investigated the subacute toxic potential of D-D. The sample used contained 25% cis-1,3-dichloropropene, 27% trans-1,3- dichloropropene, 29% 1,2-pop, and lesser amounts of 3,3-dichloropro- pene, 2,3-dichloropropene, and related chlorinated hydrocarbons. Male and female Fischer 344 rats and CD1 mice were exposed by inhalation 6 hours/day, 5 days/week for 6 or 12 weeks to 0-, 5-, 15-, or 50-ppm

Toxicity of Selected Contaminants 333 concentrations of D-D. Significant decreases in body weight gain in com- parison to controls were seen in male rats at all three exposure levels during the first 6 weeks, but not thereafter. Hematological evaluations revealed statistically significant decreases in white-blood-cell counts at 12 weeks in the 15-ppm males and the 50-ppm females, although the authors did not consider these changes to be of toxicological significance. No exposure-related effects on urinalysis or clinical chemistry parameters were observed. At 6 and 12 weeks, increases were noted in liver-to-body- weight ratios of male rats and in kidney-to-body-weight ratios of female rats at the 50-ppm exposure level. No histopathological changes were seen in rats, but diffuse hepatocytic enlargement was observed in the 50-ppm male and female mice at 12 weeks. Chronic Elects Chronic toxicity data on 1,2-pop were obtained in an NTP (1983a) carcinogenicity bioassay. Female Fischer 344/N rats and B6C3F~ mice of each sex were given reagent-grade 1,2-pop in corn oil by gavage 5 days/week for 103 weeks in doses of 0, 125, or 250 mg/kg. Male Fischer 344/N rats similarly received doses of 0, 62, or 125 mg/kg. A dose-related decrease in body weight gain was observed in both male and female rats throughout most of the study. There was a significant reduction in survival of the high-dose female rats. The high-dose female rats also experienced an increased incidence of focal and centrilobular necrosis of the liver. Hepatocytomegaly, as well as focal and centrilobular hepatic necrosis, were seen in male mice receiving 1,2-pop. The inci- dence of these lesions appeared to be dose dependent, although it was not clear whether their occurrence in the low-dose group was frequent enough to be statistically significant. The chronic oral toxicity of 1,3-dichloropropene was evaluated during the course of an NTP (1985a) carcinogenicity bioassay of Telone II, a soil fumigant containing approximately 42% cis-1,3-dichloropropene, 46% trans-1,3-dichloropropene, 2.5% 1,2-dichloropropane, 1.5% of a trichlo- ropropane isomer, 1% epichlorohydrin (a known carcinogen), and the remainder other compounds. Male and female Fischer 344/N rats and B6C3F~ mice were given the formulation in corn oil by gavage 3 times weekly for 104 weeks. The rats received doses of 0, 25, or 50 mg/kg; the mice received 0, 50, or 100 mg/kg. Blood samples were taken from designated rats every 4 weeks up to 39 weeks for hematological studies and up to 69 weeks for assessment of clinical chemistry parameters. No toxicologically important changes were observed in the hematology or the clinical chemistry studies. Rats of each sex were sacrificed after 9, 16, 21, 24, and 27 months to assess the development of histopathological changes over time. Basal cell hyperplasia of the forestomach was found in male and female rats 9 to 16 months after dosing began. No other

334 DRINKING WATER AND HEALTH lesions were observed in any tissue examined during the time-course study. The following histopathological findings were reported at the terminal sacrifice: dose-related increase in basal cell hyperplasia of the forestomach of male and female rats as well as high-dose female mice; edema of the submucosa of the urinary bladder in high-dose male and female mice; dose-related increase in the incidence of epithelial hyperplasia of the uri- nary bladder in male and female mice; and increased incidences of ne- phropathy in female rats and hydronephrosis in female mice. No chronic toxicity studies of 1,2,3-TCP were found in the literature. Neurotoxicity There is no evidence that these compounds, alone or in combination, induce permanent neurological deficits (structural or func- tional). However, the CPs are similar to other short-chain aliphatic, hal- ogenated hydrocarbons in that they can cross the blood-brain barrier and induce anesthesia or other transient effects on the central nervous system (Sidorenko et al., 19761. There is a time-dose relationship with respect to the onset of ataxia and loss of righting reflex that is additive for the three compounds at all doses and combinations tested. There was no indication of synergism or antagonism from the results seen in a multi- factorial experiment. Mutagenicity Principe et al. (1981) and Carere and Morpurgo (1981) reported that 1,2-pop induced forward mutations (8-azaguanine resis- tance) in Aspergillus nidulans but did not induce forward mutation (strep- tomycin resistance) in Streptomyces coelicolor. Principe et al. (1981) found that 1,2-pop at 11 mg/plate was marginally mutagenic to Salmo- nella typhimurium TA1535 and TA100. De Lorenzo et al. (1977) found that 1,2-pop at 10, 20, or 50 mg/plate was mutagenic in the Ames Salmonella test with and without mammalian enzyme activation mixture (S91. Stolzenberg and Hine (1980) detected no mutagenic activities in Salmonella typhimurium TA100 with 1,2-pop up to 10 mg/plate. Several genotoxicity studies with 1,2-pop were conducted by NTP (1983a). The compound was not a mutagen in the Ames mutagenesis bioassay but did induce sister-chromatic exchange at levels of approximately 1 mg/ml and chromosome aberrations (unspecified) at similar levels without S9. With S9, the compound induced these changes at approximately half this level. 1,3-Dichloropropene (with 1% epichlorohydrin added as a stabilizer) induced mutations in the Ames Salmonella test strains TA1535, TA100, and TA1978 (de Lorenzo et al., 1977) and sex-linked recessive mutations in Drosophila (Climie et al., 19791. Removal of the polar metabolites from 1,3-dichloropropene eliminated the direct-acting mutagenic activity (Talcott, 1981; Talcott and King, 1984~.

Toxicity of Selected Contaminants 335 Commercial 1,3-dichloropropene preparations were mutagenic in the Ames assay without microsomal extract S9 (Stolzenberg and Hine, 1980~. However, Talcott and King (1984) found that pure 1,3-dichloropropene isomers were not mutagenic in an Ames test in the absence of S9. They showed that oxygenated and chlorinated degradation products of 1,3- dichloropropene are the direct-acting mutagens in an Ames mutation assay. Stolzenberg and Hine (1980) reported that 1,2,3-TCP is mutagenic to Ames Salmonella typhimurium TA100 at a concentration of 0.1 Amos/ plate but only in the presence of S9 microsomal extract. They concluded that 1,2,3-TCP is mutagenic after metabolic activation. 1,2,3-TCP was not tested for its ability to include the SOS (error-prone DNA repair) response in Ames Salmonella tester strains. However, a chemically similar compound (1,2-dibromo-3-chloropropane) was tested and was found to be very mutagenic by that assay (Ohta et al., 19841. Carcinogenicity Heppel et al. (1948) exposed C3H mice to 400-ppm concentrations of 1,2-pop 37 times for 4 to 7 hours each exposure. Most of the mice died during the course of the experiment. Nine that died after having received 14 to 28 exposures had moderate to marked congestion and fatty degeneration of the liver, seven had extensive centrilobular necrosis of the liver, and six of eight had slight to moderate fatty degen- eration of the kidney. Only three mice survived both the exposures and a subsequent observation period of 7 months, by which time they were 13 months of age. These animals had developed multiple hepatomas. NTP (1983a) conducted a carcinogenesis bioassay for 1,2-pop (>99% pure). The chemical was administered in corn oil by gavage to female Fischer 344/N rats at doses of 125 or 250 mg/kg bw, to male and female B6C3F~ mice at doses of 125 or 250 mg/kg bw, and to male Fischer 344/ N rats at doses of 62 or 125 mg/kg bw. Doses were administered 5 days/ week for 103 weeks. There was no evidence of carcinogenicity for male Fischer 344/N rats. For female rats there was equivocal evidence of car- cinogenicity in that 250-mg/kg doses of 1,2-pop caused a marginally increased incidence of adenocarcinomas in the mammary gland. These borderline malignant lesions occurred concurrently with decreased survival and reduced body weight gain. The increased incidence of hepatocellular adenomas provided some evidence that 1,2-pop was carcinogenic for male and female B6C3F~ mice. No studies have been conducted with the compound added to drinking water. Telone II was given in co~n oil by gavage 3 days/week for 104 weeks at doses of 0.25 or 50 mg/kg to 52 male and 52 female Fischer 344/N rats and at doses of 0.50 or 100 mg/kg to 50 male and 50 female B6C3F~ mice. These exposures induced time- and dose-dependent tumors of the forestomach and liver nodules in male Fischer 344/N rats (NTP, 1985a).

336 DRINKING WATER AND HEALTH TABLE 9-12 Tumor Incidence in Mice Gavaged with Technical-Grade 1,3-Dichloropropene (Telone II)ab Tumor Dose Tumor Animal Sex Site (mg/kg/day) Rates B6C3F~ mouse Female Urinary bladder 0 0/50 50 8/50 100 21/48 aBased on data from NTP, 1985a. bContains 1% epichlorohydr~n, 2.5% 1,2-pop, and 1.5% trichloropropene. There was equivocal evidence for tumor induction in the forestomach of female Fischer 344/N rats. In female B6C3F~ mice, the compound induced tumors at several sites, including the urinary bladder, forestomach, and the lung (adenoma). Based on the evidence presented by NTP (1985a), Telone II is a carcinogen in Fischer 344/N rats and B6C3F~ mice. Van Duuren et al. (1979) reported that subcutaneous injection of the cis isomer of 1,3-dichloropropene produces sarcomas in Swiss mice but that tumors were not induced by repeated skin applications. cis-1,3-Di- chloropropene failed to elicit a response in Swiss mice in an initiation- promotion assay using phorbol myristate acetate (PMA) as a tumor pro- moter (van Duuren et al., 19791. These data suggest that 1,3-dichloropropene may be carcinogenic in mice. Because the data are incomplete, however, and because there is evidence that contaminants of the preparations may be more mutagenic than 1,3-dichloropropene itself (Talcott and King, 1984), it is not possible to determine the organ specificity or dose response of the effects. No information was found pertaining to the carcinogenicity of 1,2,3- TCP. However, because of its structural similarities to certain chemicals that have been found to be carcinogenic in animals (e.g., 1,2-pop, 1,3- dichloropropene, ethylene dibromide, and 1,2-dibromo-3-chloropropane), it would be prudent to limit exposure to it. There are insufficient data on the acute, subacute, and chronic noncarcinogenic toxicity of 1,2,3-TCP to conduct a toxicity risk assessment. The committee recommends that short- and long-term oral toxicity studies and a carcinogenicity bioassay be conducted with 1,2,3-TCP. Carcinogenic Risk Estimate The results of the NTP (1985a) study in which Telone II was found to induce tumors in rats and mice are used to estimate carcinogenic risk. The tumor incidence rates in Table 9-12 were used to make statistical estimates of lifetime risk. An upper 95% confi- dence bound for human exposure is shown in Table 9-13.

Toxicity of Selected Contaminants TABLE 9-13 Carcinogenic Risk Estimates for Technical-Grade 1,3-Dichloropropene (Telone II)a from Generalized Multistage Modelb Upper 95% Confidence Estimated Human Estimate of Lifetime Animal Sex Lifetime RiskC Cancer RiskC B6C3F~ mouse Female 0.5 x 10-6 1.1 x 10-6 aContains 1% epichlorohydrin, 2.5% 1,2-pop, and 1.5% tuchloropropene. bBased on data from NIP, 1985a. CAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ,ug/liter. Developmental Effects Data are not available. Reproductive Elects Osterloh et al. (1983) determined that 1,3-di- chloropropene administered by intraperitoneal injections of 0 to 75 ma/ kg/day for 5 days had no effect on sperm count, sperm morphology, or testicular weights in mice. However, several substances that are known to induce sperm abnormalities in humans have not been observed to pro- duce positive results in the mouse assay because of species differences or because observations were made at different times. In the study by Osterloh and colleagues, observations wer_ made on the 35th day after exposure; thus, earlier or later effects may have been missed. The authors noted, therefore, that care must be taken when interpreting this negative response with 1,3-dichloropropene. Hardin et al. (1981) administered intraperitoneal 37-mg/kg bw doses of 1,2,3-TCP in corn oil to Sprague-Dawley rats. This dose caused ma- ternal toxicity but did not produce fetal toxicity or teratogenesis. CONCLUSIONS AND RECOMMENDATIONS The toxicity of the chloropropanes and chloropropenes has been ex- amined in only a few studies. Most of the studies reported in the literature have focused on the toxicity of two or more of these compounds in mixtures. 1,2-pop causes injurer in the liver, kidneys, and adrenals following acute, subacute, and chronic exposure. 1,2,3-TCP produces the same target organ toxicity under similar exposure conditions. Following long- term gavage studies, 1,3-dichloropropene (Telone II) was found-to be carcinogenic in rats and mice. The toxicity of these substances should be reassessed when additional information becomes available.

338 DRINKING WATER AND HEALTH Dl(2-ETHYLHEXYL) PHTHALATE (DEHP) bis(2-Ethy~hexyI) I,2-benzenedicarboxylate CAS No. ~ 17-~-7 RTECS No. Tt0350000 COOCH2 CH(C2 Hs )(CH2 )3 CH3 ^< COOCH2 CH(C2 He )(CH2 )3 CH3 MONO(2- ETHYLH EXYL) PHTHALATE (M EN P) Mono(2-ethylhexyl) 1,2-benzenedicarboxylate CAS No. 4376-20-9 RTECS No. TI2500000 COOCH2 CH(C2 He )(CH2 )3 CH3 ~,COOH Di(2-ethylhexyl) phthalate (DEHP) can be hydrolyzed in the environ- ment to produce mono(2-ethylhexyl) phthalate (MEHP) and 2-ethylhex- anol (2-EH). Pharmacokinetic studies indicate rapid conversion of DEHP in the body to MEHP and 2-EH so that plasma levels of MEHP are much higher than those of DEHP. Thus, DEHP and MEHP are considered together in this analysis. DEHP is widely used as a plasticizer to impart flexibility in many consumer plastic products and medical devices. Some formulations of polyvinyl chloride pipe used to carry drinking water contain as much as 30% of the compound. An estimated 2 billion pounds of DEHP are pro- duced annually worldwide, making DEHP one of the most widespread synthetic environmental chemicals (Thomas and Thomas, 19841. DEHP has been found in air over ocean water at levels ranging from 0.25 ppt to 1.8 ppt (IARC, 19821. Concentrations in water have varied widely: 5 ng/liter in the North Atlantic, 80 ng/liter in the Gulf of Mexico, and 130 ng/liter near the Gulf Coast. Ocean sediments in the Gulf of Mexico contained 2 ng/g, compared with 69 ng/g in the Mississippi delta. In surface water near chemical plants, DEHP levels of 1 to 50 ~g/liter have been detected. In a study of the environmental transport of DEHP, there was an initial concentration of 200 ~g/liter in a particular industrial

Toxicity of Selected Contaminants 339 effluent, decreasing concentrations through various water treatment steps, and a final level of 0.6 ~g/liter in finished drinking water in Philadelphia (Sheldon and Hites, 19791. DEHP is found in animals and animal products in the human food chain. The highest levels have been detected in milk (31.4 mg/kg, fat basis) and cheese (35 mg/kg, fat basis). These data indicate that DEHP can accu- mulate in animal tissue. This is further established by an 886-fold con- centration found in fathead minnows exposed to 4.6-,ug/liter concentrations of ~4C-labeled DEHP (Mehrle and Mayer, 19761. DEHP is metabolized by several species of bacteria (Engelhardt et al., 19751. Because of the wide distribution of DEHP in the environment, the rates of DEHP degradation by enzymatic and nonenzymatic reactions in different environmental compartments may vary widely. Thus, an overall environmental half-life may not be useful in estimating the residue levels in particular locations. Occasional acute exposures to very high levels may result when DEHP leaches from plastics used in medical apparatus such as storage bags and tubing for blood transfusions. Blood plasma stored in polyvinyl chloride bags has had DEHP levels of up to 250 mg/liter (Pik et al., 1979~. METABOLISM DEHP appears to be efficiently absorbed from the gut of the rat, as indicated by urinary excretion of more than 90% of the activity in an oral 2,000-mg/kg dose of ~4C-labeled DEHP (Williams and Blanchfield, 19741. Very little of the diester is absorbed intact: DEHP is hydrolyzed by in- testinal esterases, and the product MEHP is the primary form absorbed (IARC, 1982; Oishi and Hiraga, 1982; White et al., 19801. Comparative studies with rats, dogs, and miniature pigs have shown major differences in absorption rates and excretion patterns (Ikeda et al., 1980~. Absorption, metabolism, and excretion of ~4C-labeled DEHP were determined in an- imals that had been given 50-mg/kg bw oral doses of DEHP for 21 to 28 days. Approximately 84% of the radioactivity was excreted in the urine and feces of rats during the first 24 hours; during this time, excretion in dogs and pigs was 67% and 37%, respectively. Fecal excretion predom- inated in dogs, whereas urine was the major route of excretion in rats and pigs. Excretion was virtually complete in all species within 4 days. Blood concentrations of DEHP and MEHP in rats reached maximum levels-within 6 to 24 hours of administration. In the heart and lungs, however, the highest levels were reached within 1 hour (Oishi and Hiraga, 19821. At 6 hours after administration, the highest ratios of M1SHP to DEHP (mol%) were found in the testes (210%~; other tissues contained less than 100%. MEHP disappeared exponentially with half-life values

340 DRINKING WATER AND HEATH ranging from 23 to 68 hours for the different tissues; the half-life of DEHP ranged from ~ to 156 hours. Values for several tissues were slightly longer for MEHP than for DEHP; e.g., the values for MEHP and DEHP in blood were 23.8 and 18.6 hours, respectively. There is little or no accumulation of radioactivity in rats following repeated dietary treatment. The highest concentrations were found in liver (110 to 165 mg/kg) and adipose tissue (60 to 80 mg/kg) after subchronic administration of 5,000 mg/kg in the diet. Estimated half-lives for DEHP and its metabolites in rats are 3 to 5 days for fat and 1 to 2 days for other tissues (Daniel and Bratt, 19741. Rats have been reported to metabolize DEHP to 5-keto-2-ethylhexyl phthalate, 5-carboxyl-2-ethylpentyl phthalate, 5-hydroxy-2-ethylhexyl phthalate, and 2-carboxymethylbutyl phthalate after initial hydrolysis to MEHP (Albro and Moore, 1974; Albro et al., 1982; Daniel and Bratt, 1974; IARC, 1982~. Only a small amount (less than 5%) is completely hydrolyzed to phthalic acid. In contrast to rats, African green monkeys (Albro et al., 1981) and ferrets (Lake et al., 1976) excrete DEHP metab- olites in urine as glucuronide derivatives of MEHP. Apparent species differences in DEHP metabolism could be due to differences in the in- testinal microflora. HEALTH ASPECTS Observations in Humans Ganning et al. (1984) examined dialysis patients for liver changes that may result from their treatment regimen. They estimated that these patients were receiving approximately 150 mg of DEHP per week intravenously during their treatment. At 1 month, no morphological liver changes were observed by liver biopsy. At 1 year, peroxisomes were described as being "significantly higher in number." No other observations were made. It was assumed by these investigators that the production of liver peroxi- somes was the result of the patients' exposure to DEHP. However, the livers in dialysis patients compared with those of healthy individuals would be exposed to a greater number of blood contaminants at higher levels because of their impaired clearance abilities. There were no other data describing the effects of DEHP in humans other than those proposed as possible effects in persons receiving intra- venous solutions. Observations in Other Species Acute Effects Several studies, beginning as early as 1943 (Hodge, 1943), have been conducted to examine the acute toxic effects of DEHP.

Toxicity of Selected Contaminants 341 In these early studies and in subsequent investigations (Thomas et al., 1978), DEHP was found to have low acute toxicity. The single oral LDsoS for DEHP vary for other routes of exposure, partly because of its poor solubility in biological solutions. The principal acute effects reported were mild irritation of the gastrointestinal tract or of other sites of administra- tion. Chronic Effects Exposures of Sherman rats to 4,000-ppm (300-mg/ kg bw) dietary concentrations of DEHP for 2 years induced no tumors (Carpenter et al., 19531. Nor were any tumors observed in Wistar rats given DEHP at 5,000 ppm (375 mg/kg bw) in the diet for 2 years (Harris et al., 19561. The information from these studies was limited because of the small sample size or limited extent of pathological examination. No tumors were reported in rats treated for 2 years with 20,000 ppm DEHP in the diet (Ganning et al., 1984~. A bioassay more in keeping with current standards was conducted by the National Toxicology Program (NIP, 1983b). In this study, Fischer 344 rats of both sexes were exposed to 6,000 and 12,000 ppm in the diet and B6C3F~ mice of both sexes were exposed to 3,000 and 6,000 ppm in the diet, all for 2 years. An increased yield of hepatocellular tumors was observed in both sexes of both species. There was a trend toward increasing yield with increasing dose. Mutagenicity The committee has relied primarily on published, peer- reviewed reports of studies on mutagenicity. The International Program for Chemical Safety (IPCS) of the World Health Organization provided additional information through a large study designed to evaluate the performance of a variety of short-term in vitro assays with a series of chemicals of interest in terms of their genotoxicity as it relates to carcin- ogenic activity. Its summary of some 90 individual sets of experiments yielded a large data base on the activity of DEHP in a variety of short- term genotoxicity tests. The primary research reports reviewed by IPCS have not yet been published. Thus, they have not undergone peer review. The IPCS compared the number of positive and negative assays, pre- sented the results of each assay, and discussed the results (Ashby et al., 19851. The committee recognizes that these data should be reevaluated when the results are published in the scientific literature. Studies conducted in several laboratories to examine DEHP and its principal metabolites, MEHP and 2-EM, have failed to show mutagenic activity in a variety of bacterial mutagenesis assays (CMA, 1982; Kirby et al., 1983; Kozumbo et al., 1982; Seed, 1982; Yoshikawa et al.,^ 1983; Zeiger et al., 1982, 1985~. In the IPCS study, all five assays for muta- genesis in Salmonella were negative (Ashby et al., 19851.

342 DRINKING WATER AND HEALTH There is one report that DEHP and MEHP were mutagenic in the Ames Salmonella test and in Escherichia cold (Tomita et al., 1982~. The same investigators described a positive response in the rec-assay with Bacillus subtilis. This report is in conflict with other published observations. Fur- thermore, in the Salmonella test, MEHP was reported to be mutagenic without exogenous metabolic activation, despite the fact that MEHP almost certainly will not react with DNA. In the IPCS study, DEHP was reported to be negative in five of six assays for yeast gene mutation (Ashby et al., 1985~. DEHP, MEHP, and 2-EH were not mutagenic in the L5178Y mouse lymphoma mutagenesis assay (CMA, 1982; Kirby et al., 1983) or in the Chinese hamster ovary (CHO) mutagenesis assay (Phillips et al., 19821. In the IPCS study, DEHP was negative in five of six mammalian-cell gene mutation assays (Ashby et al., 19851. In contrast, Tomita et al. (1982) showed MEHP to be mutagenic in V79 cells and both DEHP and MEHP to induce point mutations in embryonic Syrian hamster embryo cells exposed transplacentally. No activity was seen in a mouse bone-marrow micronucleus assay with DEHP, MEHP, or 2-EH (CMA, 1982) or in a rat bone-marrow cytoge- netics assay (Putman et al., 19831. In the IPCS study, all three assays for chromosome aberrations in mammalian-cell cultures were negative (Ashby et al., 19851. Abe and Sasaki (1977) reported increased sister-chromatic exchange (SCE) induction in CHO cells after 24-hour exposure to DEHP with a dose-related increase from 8.8 to 11.0 SCE per cell over a dose range of about 4 to 400 ~g/ml. They did not observe chromosome ab- errations in either CHO or human cells. However, Tomita et al. (1982) reported that DEHP and MEHP also induced chromosome aberrations and SCEs in Chinese hamster V79 cells. No cytotoxicity information was provided. DEHP did not produce chromosome damage in human lym- phocytes or in human fetal lung cells (Stenchever et al., 19761. There is one report that MEHP but not DEHP was clastogenic at high concentrations in CHO cells (Phillips et al., 19821. The same study re- ported both compounds as negative in the CHO sister-chromatic exchange assay. These results are curious because the SCE assay is usually the more sensitive cytogenetic end point. In the Phillips et al. study, however, cells were exposed for only 2 hours, and the data reported for MEHP showed a positive trend (not statistically significant). SCEs per cell increased from 12.8 in the controls to 14.5 at the highest dose. Further studies by the same group suggest that induction of chromosome aberrations in the CHO cells was the result of severe cytotoxicity (Phillips et al., 19861. The authors found that CHO cultures were not viable in the presence of 0.5 rnM MEHP as measured by Trypan Blue exclusion, whereas the chro- mosome effects were only seen above 1.0 rnM. MEHP is a detergent

Toxicity of Selected Contaminants 343 strong enough to lyse cells, and this may be occurring at the higher concentrations. More recently, NTP-sponsored studies of DEHP in CHO cells have demonstrated a modest, statistically significant, reproducible increase in SCE after a 24-hour exposure to high doses of DEHP in the absence of S9 (Ashby et al., 1985~. In the IPCS study, all three of the assays for mutagenesis in somatic cells in Drosophila were negative (Ashby et al., 19851. DEHP, MEHP, and 2-EH did not induce DNA repair in metabolically competent primary rat hepatocyte cultures (Butterworth et al., 1984; CMA, 1982; Hodgson et al., 1982; Kornbrust et al., 19841. Similarly, no activity was seen with DEHP or MEHP in the primary human hepatocyte DNA repair assay (Butterworth et al., 19841. In the IPCS study, DEHP was negative in all three of the DNA repair assays done by autoradiography (Ashby et al., 19851. In the IPCS study, a positive result was obtained in a primary hepatocyte DNA repair assay with scintillation counting (Ashby et al., 19851. Scin- tillation counting cannot distinguish a general DNA repair response from a stimulation of cells in S-phase (Doolittle et al., 1984~. Other hypolip- idemic agents have been shown to stimulate S-phase DNA synthesis in primary hepatocyte cultures (Bier) et al., 19841. It is currently believed that cell transformation assays are capable of measuring some of the stages in the progress of a cell from the normal to the malignant state. The C3H 10T2 assay measures the conversion of cells from a preneoplastic to a tumorigenic state. The assay measures both initiating and promoting events (Frazelle et al., 19841. Both DEHP and MEHP were negative as complete transforming agents, initiators, or pro- moters in the 10T2 cell transformation assay (J. Sanchez, D. Abernethy, and C. Boreiko, Chemical Industry Institute of Toxicology, Research Triangle Park, N.C., personal communication, 1985~. DEHP, MEHP, and 2-EH were also negative in the BALB/3T3 cell transformation assay (CMA, 1982). The Syrian hamster embryo cell transformation assay measures trans- formation from the normal to the preneoplastic state. This assay is sensitive not only to chemicals that affect the primary DNA sequence but also to agents that cause aneuploidy. Several carcinogens, such as asbestos and diethylstilbestrol (DES), that are negative in traditional mutagenicity tests are positive in the Syrian hamster embryo cell transformation assay (Barrett et al., 1981; Hesterberg and Barrett, 19841. The action of these agents appears to lie in the induction of aneuploidy (Barrett et al., 19841. DEHP has been shown to be positive in the Syrian hamster embryo cell trans- formation assay at concentrations ranging from 1.0 to 100 mg/ml (0.0026 to 0.26 mM) with reasonable cell survival (Barrett and Lamb, 19851. In the IPCS study, four of five assays for cell transformation were positive

344 DRINKING WATER AND H"LTH and three of the five IPCS assays for aneuploidy were positive (Ashby et al., 1985~. There is a report that DEHP was weakly active in a hamster transpla- cental mutagenesis and transformation assay (Tomita et al., 19821. In this experiment, very high doses were administered, and the control values were much higher than expected for an assay of this type. Kornbrust et al. (1984) reported that DEHP was negative in the in vitro Chinese hamster V79 metabolic cooperation assay for tumor promoters (upper concentration of 0.12 ~g/ml). This is in contrast to the findings of Malcolm et al. (1983), who reported that DEHP inhibited metabolic cooperation in the same assay at concentrations of 1 to 30 ~g/ml. Conclusions on Mutagenicity DEHP/MEHP have been subjected to extensive testing in short-term tests. Despite two positive in vitro tests for bacterial and mammalian cell mutagenicity, the committee concluded from the weight of the evidence that DEHP and MEHP were not mutagenic for bacterial or mammalian cell systems, with or without metabolic activation. DEHP is not mutagenic in Drosophila, does not induce DNA repair in metabolically competent cultures of human or rat hepatocytes, and is largely negative for the induction of chromosome aberrations. However, it has been reported to be a weak inducer of SCEs. DEHP is able to induce aneuploidy in cells in culture and is capable of inducing cell transformation as shown in the Syrian hamster embryo. However, there are reasons to be cautious about concluding that the sole mechanism of action of DEHP is the induction of aneuploidy or other chromosomal events, such as sister-chromatic exchanges. First, other biological effects exerted by DEHP in the whole animal may be as im- portant as or more important than its carcinogenic activity. Second, al- though reported to be a weak inducer of SCEs, DEHP does not induce other chromosome abnormalities associated with agents such as benzene or DES, which induce other genetic activities such as clastogenicity and polyploidy (Ashby et al., 19851. Possible Mechanisms of Carcinogenicity According to several re- searchers, peroxisome proliferators such as DEHP produce secondary ge- netic toxicity by stimulating biosynthesis of liver-cell peroxisomes, increasing peroxisomal hydrogen peroxide-generating oxidases (Cerutti, 1985; War- ren et al., 1982), and modifying the pattern of enzyme activities. This results in an excessive production of hydrogen peroxide (paralleled by a decrease in catalase) and, by further reactions, of hydroxy ~ OH) radicals. Hydrogen peroxide, hydroxy radicals, and other forms of^active oxygen can cause reactions that result in mutations, sister-chromatic exchanges, chromosome aberrations, and cancer (Cerutti, 19851. Active oxygen

Toxicity of Selected Contaminants 345 appears to play a role mostly in the promotion phase but also in the progression phase of carcinogenesis (e.g., by inducing chromosome ab- errations). According to Cerutti (1985), the metabolite MEHP may also produce a prooxidant state by inhibiting the electron transport chain. DEHP induces several functional changes in the liver, including he- patomegaly and peroxisomal proliferation (Miyazawa et al., 1980) in a manner similar to that of hypolipidemic drugs such as clofibrate. Livers from rats fed a diet containing 2% DEHP had an increased number of peroxisomes after a few days. Liver biopsies of patients undergoing di- alysis for one year have an increase in peroxisomes, possibly related to DEHP exposure (Ganning et al., 1984~. There are conflicting data on the role of peroxisome proliferation in DEHP's carcinogenicity. Reddy and Lalwani (1983) hypothesized that peroxisomal proliferating agents may constitute a novel class of carcin- ogens. Warren et al. (1982) proposed that following peroxisomal prolif- eration, genotoxic activity results from an increase in the production of DNA-damaging reactive oxygen species. Accordingly, the ability of DEHP to induce DNA damage or repair has been examined in the in viva he- patocyte DNA repair assay (Mirsalis and Butterworth, 19801. DNA dam- age has also been measured by alkaline elusion of cellular DNA from the same cultures (Bermudez et al., 19821. Female rats were treated with 12, 000-ppm concentrations of DEHP in the diet for 30 days or for 30 days followed by a 500-mg/kg dose of DEHP by Savage for 14 days. Male rats were treated by Savage with 500-mg/kg doses of DEHP 2, 12, 24, or 48 hours before sacrifice or with 150 mg/kg by gavage for 14 days. No chemically induced DNA damage or repair was seen under any of the conditions of the study (Butterworth et al., 1984~. In a similar study, male Sprague-Dawley rats were given a single oral DEHP dose of 5,000 mg/kg bw or fed a diet containing 20,000-ppm concentrations of DEHP for up to 8 weeks. Genotoxicity was evaluated in the in viva hepatocyte DNA repair assay. No chemically induced DNA repair was observed, even in animals pretreated with 3-amino-lH-1,2,4,- triazole to inhibit catalase activity, thus maintaining any elevated peroxide levels that might have existed (Kornbrust et al., 19841. These studies indicated that neither the parent compound nor its metabolites bound to the DNA to elicit the DNA repair response. Similarly, Lutz (1986) ob- served no covalent binding to the DNA in female Fischer 344 rats given DEHP by oral Savage in olive oil (the covalent binding index was estimated to be less than 0.051. From these studies, the committee concluded that neither DEHP nor its metabolites bind to the DNA in the whole animal. More recent evidence challenges the hypothesis that peroxisomal pro- liferation is the basis of the carcinogenic activity of DEHP. With the alkaline elusion assay, no significant DNA strand breaks were detected in

346 DRINKING WATER AND HEALTH the cells from the treated animals (Butterworth et al., 19841. Administra- tion of DEHP in concentrations ranging from 500 to 5,000 ppm in the diet of male Sprague-Dawley rats caused large increases in carnitine palmitoyltransferase, carnitine acetyltransferase, and Q-oxidation capacity (Morton, 19791. Similar increases have been observed in Fischer 344 rats (Butterworth et al., 19841. Therefore, it is likely that the animals fed 6,000-ppm concentrations of DEHP in the long-term bioassay must have had significant peroxisomal proliferation. Neither the female nor the male rats developed a statistically significant increase in hepatocellular carci- nomas at that dose, although a nonsignificant increase was seen in female rats. Similarly, di(2-ethylhexyl) adipate (DEHA) is a peroxisome proliferator in rodents (Ready and Lalwani, 1983) but produced no tumors in Fischer 344 rats given 25,000-ppm concentrations of DEHA in the diet for 2 years (NIP, 1982c). However, DEHA did produce tumors in mice at the same dose. Furthermore, DEHP is more effective in inducing peroxisomal en- zymes in male than in female Sprague-Dawley (Osumi and Hashimoto, 1978) or Fischer 344 rats (CMA, 1982; F. E. Mitchell et al., 19841. In contrast, the female rats had a greater increase in the rate of liver tumors than did the male rats (NIP, 1983b). Lipid peroxidation measured in animals given 20,000-ppm concentrations of DEHP in the diet for 6 weeks plus a single 5-g/kg bw DEHP dose by gavage did not differ from controls (Kornbrust et al., 19841. Thus, although a role for peroxisomal proliferation in carcinogenesis cannot be ruled out, genotoxic activity does not appear to be associated with DEHP-induced peroxisomal proliferation at the level of sensitivity that has been measured, suggesting that peroxisomal proliferation alone is not sufficient to produce tumors. More research should be done to test this hypothesis. Carcinogenicity Promotional activity is a plausible hypothesis for the mechanism of action of DEHP in light of its lack of mutagenic activity in most cell culture and animal models. There are several models for initiation and promotion in the rodent liver in which the animals are treated with a single dose of initiator followed by an extended period of treatment with promoter. The end points generally observed are focal hepatocellular proliferative lesions (FHPLs) such as foci positive for ~y-glutamyl trans- peptidase (GGT). Although it has never rigorously been shown that GGT- positive foci are actual precursors of liver tumors, the evidence indicates that these assays reflect initiating and promoting events. Popp et al. (1985) reported no increase in enzyme-altered foci, preneoplastic nodules, or liver tumors in female Fischer 344 rats under either of two regimens: DEHP given as an initiator followed by a growth selection regimen to

Toxicity of Selected Contaminants 347 express initiated sites or diethylnitrosamine (DEN) given as the initiator followed by 12,000-ppm dietary concentrations of DEHP for 6 months as a promoter. Similarly, no promoting activity of DEHP was seen in female Fischer 344/NCr rats initiated with DEN followed by 14 weeks of pro- motion with 12,000-ppm concentrations of DEHP (Ward et al., 19861. As with the rats, DEHP was found to have no initiating activity in B6C3F~ mice (Ward et al., 1983~. In contrast to the results found in the rats, DEHP given in the diet in concentrations ranging from 3,000 to 12,000 ppm was shown to promote FHPL in mouse liver (Ward et al., 1983, 1984a). Phenobarbital is a standard positive control with known promoting activity used in these assays. Whereas phenobarbital induced an increase in both the number and size of the FHPLs, the effect of DEHP was to increase FHPL size but not number (Ward et al., 19831. Pheno- barbital required 168 days of continuous exposure for a promoting effect to be evident. DEHP was an effective promoter after 28 days (Ward et al., 1984a). In skin-painting studies, DEHP displayed weak complete promoting activity but did show second-stage promoting activity in Sencar mouse skin. In contrast, the compound was inactive in CD1 mouse skin-painting experiments (Ward et al., 19861. Thus, DEHP does appear to have species- specific promoting activity in the mouse. This is consistent with the cancer studies, which yielded a response in the mouse liver at lower doses than in the rat liver (NTP, 1983b). Stott et al. (1981) suggested that some carcinogens are not directly genotoxic but may exert their effects through cell turnover activity, either by producing cytotoxicity, thus resulting in increased regenerative DNA synthesis, or by inducing hyperplasia directly (Stott et al., 19811. Con- tinual excessive cell turnover could result in an increased frequency of mutations and in promotional effects. One observed action of DEHP on the DNA of Fischer 344 rats was a small increase in the number of hepatocytes in the S-phase (Butterworth et al., 19841. This is consistent with the increased liver-to-body-weight ratios observed upon prolonged administration of DEHP (F. E. Mitchell et al., 1984; Miyazawa et al., 1980; Morton, 19791. Similar hepatomegaly and mitogenic stimulation has resulted from the carcinogenic hypolipidemic peroxisomal proliferators (Moody et al., 1977; Reddy et al., 1976, 19781. The spontaneous and DEHP-induced tumor incidences in the mice were greater than those in the rats (NTP, 1983b). These responses are similar to those of a variety of other compounds that induce tumors more easily in mice than in rats and that appear to be without mutagenic activity (Doull et al., 1983; Ward et al., 19791. If the carcinogenic activity of DEHP were due to effects of cell turnover, one might expect mice to be more susceptible to induction of cell repli-

348 DRINKING WATER AND H"LTH TABLE 9-14 Hepatocellular Carcinoma Incidence in Rats and Mice Fed DEHP for 2 Yearsa Concentration in Feed (ppm) Animal Sex Controls 3,000 6,000 12,000 Fischer 344 rat Male 1/50 1/49 5/49 Female 0/50 2/49 8/50 B6C3F~ mouse Male 9/50 14/48 19/50 Female 0/50 7/50 17/50 aBased on data from NIP, 1983b. cation. Studies show that DEHP induces an S-phase response in mouse hepatocytes that is approximately an order of magnitude greater than that seen in rats following a single dose or following the administration of diets containing 6,000 ppm for 2 weeks (Smith-Oliver et al., 19851. More work should be done in both the rat and mouse to learn how long the increased S-phase response is maintained in response to DEHP feeding. At this time, the carcinogenic effects of DEHP appear to be correlated with cell replication (NTP, 1983b). More studies should be conducted to examine the duration and shape of the dose-response curve for induced cell turnover. This is not to suggest that induction of cell replication is likely to be the sole explanation of the carcinogenic effect of DEHP, but it may be an important contributing factor. Statistically significant increases in hepatocellular carcinomas were ob- served in the NTP bioassay described above (NTP, 1983b) for female rats receiving 12,000-ppm, male mice receiving 6,000-ppm, and female mice receiving 6,000- and 3,000-ppm concentrations of DEHP in feed. The incidence of hepatocellular carcinomas in this bioassay is given in Table 9-14. Statistically significant decreases in the incidence of pituitary tu- mors, thyroid C-cell carcinomas, and testicular interstitial tumors were found among the male rats. DEHP also produced a statistically significant increased incidence of either hepatocellular carcinomas or neoplastic nod- ules in male rats (controls, 3/50; 6,000 ppm, 6/49; 12,000 ppm, 12/49~. Carcinogenic Risk Estimate In a study by the National Toxicology Program, an increase in hepatocellular tumors was observed in both sexes of Fischer 344 rats and B6C3F~ mice (NTP, 1983b). There was a trend toward increases in the number of tumors with increasing doses. The tumor incidences for the mice, the more sensitive species, are summarized in Table 9-15. The data for each sex were used to estimate a lifetime risk and an upper 95% confidence estimate of lifetime risk in humans weighing 70 kg fol-

Toxicity of Selected Contaminants 349 TABLE 9-15 Tumor Incidence in Mice Fed Di(2-ethylhexyl) Phthalatea Tumor Dose,b TumorC Animal Sex Site ppm (mg/kg/day) Rates B6C3F~ mouse Male Liver 0 14/50 3,000 (3 12) 25/48 6,000 (615) 29/50 aBased on data from NIP, 1983b. bOf the DEHP metabolite mono(2-ethylhexyl) phthalate. CHepatocellular carcinomas or adenomas. lowing a daily consumption of 2 liters of water containing the compound in a concentration of 1 ~g/liter. No estimates were calculated for children weighing 10 kg and consuming 1 liter of water; however, the risk estimates for children would be higher than those shown below due to increased consumption per body weight. The estimate of lifetime risk is based on the generalized multistage model for carcinogenesis described earlier in Chapter 8 (see Table 9-161. Calculations based on the generalized mul- tistage model indicate that a DEHP dose of 0.032 mg/kg/day presents an estimated lifetime risk to humans of 1 x 10-6. Developmental Ejects Because of the apparent ubiquity of the phthal- ate esters in the environment, their teratogenicity and reproductive toxicity, particularly of the most commonly used ester DEHP, have been investi- gated in numerous laboratories. Examination of these effects of DEHP and of its principal metabolite MEHP (Albro et al., 1973; Schulz and Rubin, 1973) have centered on the use of DEHP as a plasticizer for polyvinyl chloride polymers used in the manufacture of venous medical products and blood storage devices. DEHP has induced malformations in both rats and mice at very high dosages; more equivocal results were obtained at lower concentrations. TABLE 9-16 Carcinogenic Risk for DEHPa Estimated with the Generalized Multistage Modelb Upper 95% Confidence Estimated Human Estimate of Lifetime Animal Sex Lifetime RiskC Cancer RiskC B6C3F~ mouse Male 1.2 x 10-7 2.1 x 10-7 aBased on data from NTP, 1983b. bThe Weibull model produces similar results. CAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ,u~g/liter.

350 DRINKING WATER AND HEALTH Singh et al. (1972) administered DEHP intraperitoneally to pregnant rats at doses of 0.3 to 10.0 ml/kg on days 5, 10, and 15 of gestation. They reported increases in gross abnormalities (primarily hemangiomas) and fetal resorptions and decreased fetal weight at only the high dose tested (10.0 ml/kg, equivalent to approximately 9,810 mg/kg). Lower dosages (up to 4~905 make) had no effect. Subsequently, Singh et al. (1974) described both mutagenic and antifertility effects in male mice that re- ceived a single intraperitoneal injection of 12,800-, 19,100-, or 25,600- mg/kg doses of DEHP prior to mating with untreated females. At the highest dose, females mated with treated males over a 12-week period exhibited a decrease in the incidence of pregnancy and in numbers of implantations and offspring as well as an increase in the incidence of early fetal death. No gross or skeletal abnormalities were described. There were no effects at lower dosages. The significance of these studies (Singh et al., 1972, 1974) is reduced by the very high dosages used and the use of the relatively inappropriate intraperitoneal route of administration. In a similar study (Domon et al., 1984), males of two different strains of mouse received subcutaneous injections of DEHP (approximately 625, 2,000, and 4,900 mg/kglday) on 3 days during the first 2 weeks of the study and were then mated to females of the same strain over an 8-week period. DEHP did not induce significant dominant lethality in either strain and produced little, if any, infertility at the dosages examined. Nikonorow et al. (1973) treated female rats orally with 340 or 1,700 mg/kg/day throughout gestation and recorded a decrease in fetal body weight and an increased number of resorptions, but no gross abnormalities. Yagi et al. (1980) examined the teratogenic potential of DEHP and MEHP in mice by giving them oral doses of these compounds. Treatment with 5.0-ml/kg (4,905-mg/kg) or 10.0-ml/kg (9,810-mg/kg) concentrations of DEHP on day 7 of gestation resulted in 100% fetal death. Survival rates of fetuses exceeded 90% when the treatment was given on day 9 or 10, however. These investigators recorded both gross and skeletal abnormal- ities in live fetuses after oral treatment with 2.5 or 7.5 ml/kg (2,450 or 7,360 mg/kg) on day 7 or 8 of gestation. Administration of 70, 190, 400, 830, and 2,200 mg/kg/day in the diet of female mice resulted in decreased maternal weight gain (probably due to reduced numbers of viable fetuses) and increased resorption rate (Shiota et al., 19801. In addition, there were no viable fetuses in the two highest dose groups and a dose-related decrease in fetal weight in the 400- and 830-mg/kg groups. In this latter group, there was also a slight but sig- nificant increase in fetal malformations, primarily neural tube defects (exencephaly and spine bifida). Recent whole-body autoradiography stud- ies of the distribution of ~4C-labeled DEHP (Lindgren et al., 1982) have shown a pronounced concentration of DEHP in the neuroepithelium of x--r -- 7- -- ~— - ,~ ~ ~,

Toxicity of Selected Contaminants 35 ~ the developing embryo, which may correlate with the DEHP-induced neural tube defects. Shiota et al. (1980) have determined that the maximum no-effect level is 70 mg/kg/day, perhaps as much as 2,000 times higher than the estimated current intake by humans from environmental sources (Nakamura et al., 19791. Ruddick et al. (1981) administered 225-, 450-, and 900-mg/kg doses of MEHP to female rats by intubation on days 6 to 15 of gestation. They reported decreases in the number of litters and in the mean litter weight in the 450-mg/kg group but no skeletal or visceral abnormalities in living fetuses. Maternal toxicity at the higher dosage prevented analysis (11 of 15 treated females died). No fetal effects were recorded at the lower dosage tested, and the no-effect dosage for dams was determined to be 50 mg/kg. Tomita et al. (1982) administered a single oral dose of DEHP (0.2% to 100% of the acute LDso of 30 mg/kg) to pregnant female mice on day 6, 7, 8, 9, or 10 of gestation. Decreases in both number and body weight of fetuses and an increased incidence of gross and skeletal fetal malfor- mations were observed in dams treated with the higher dosages. The fetal LDso of DEHP was found to be 592 mg/kg, and the maximum nonlethal dose for the fetuses was calculated to be about 64 mg/kg. Tomita et al. (1982) also estimated that the ineffective maximum dosages (no-observed- effect levels) for gross and skeletal abnormalities were probably less than 789 and 670 mg/kg, respectively. In humans, the maximum intake of DEHP by blood transfusion or hemodialysis is approximately 4 mg/kg bw (Jaeger and Rubin, 19721. In an interesting comparative study, WoLkowski-Tyl et al. (1983a) re- ported teratogenic effects in mice at high doses of DEHP (0.10% and 0.15% in the feed during the entire gestation period), which induced maternal toxicity, but no such effects at doses not toxic to the dam (0.05%~. In contrast, DEHP was not teratogenic in rats at any concentration tested with the identical protocol (Wolkowski-Tyl et al., 1983b). These studies demonstrate that the effects of DEHP on fetal development and viability appear only at dosages much higher than those possible in humans (see review by Thomas et al., 1978~. Several studies have been conducted using more clinically relevant dosage regimens. Garvin et al. (1976) examined the teratogenic potential of plasma-soluble extracts of polyvinyl chloride containing approximately 1- and 3.7-mg/kg doses of DEHP administered intravenously on gestation days 6 to 15. No embryo- toxic or teratogenic effects were detected. Lewandowski et al. (1980) administered mean daily intravenous 1.3- to 5.3-mg/kg doses of similar plasma-soluble extracts of DEHP-plasticized polyvinyl chloride to rats on gestation days 6 to 15. The high doses are similar to those predicted to be received by a 60-kg human undergoing an exchange transfusion with

352 DRINKING WATER AND H"LTH 21-day-old blood. There were no significant differences in fetal weights, in the number of live and resorbed fetuses, or in the incidence of gross external, skeletal, and visceral defects between control and treated groups. The authors concluded that the tested plasma extracts were not teratogenic when administered intravenously to pregnant rats. Studies of pregnant rabbits (Thomas et al., 1979) confirmed that exposure to more clinically relevant concentrations of MEHP (1.14, 5.69, and 11.38 mg/kg daily on gestation days 6 to 19) were neither teratogenic nor embryotoxic. It is clear that only very high dosages of DEHP (or MEHP) are em- bryotoxic or teratogenic in the species tested. Levels that produce such effects generally are far greater than those anticipated for human exposure (Shiota et al., 19801. Reproductive Elects Gonadal toxicity of DEHP was first reported in the mid-1940s, when Shaffer et al. (1945) described the results of a 90- day feeding study in which male rats received dietary concentrations of 0.075%, 0.75%, 1.5%, and 5.0% DEHP in feed. These investigators reported atrophy of the seminiferous tubules and testicular degeneration resembling "senile changes" at the two highest dosages. In a two- generation feeding study (Carpenter et al., 1953) in which rats were main- tained on diets containing 0.04% to 0.4% DEHP, a decrease in the number of litters per female was reported in Fit rats in the highest dose group. Subsequent metabolic studies with the phthalate diesters have revealed that the testicular effects are mediated via the corresponding monoesters (MEHP in the case of DEHP) that are formed by partial hydrolysis of the parent compound in the gastrointestinal tract (Albro et al., 1973; Rowland et al., 19771. The alcohols also formed by this hydrolysis (2-ethylhexanol in the case of DEHP) have more recently been shown to have no activity in induction of the testicular lesions produced by the phthalate diesters (Gray and Beamand, 1984; Rhodes et al., 19841. Other investigators have described the morphology of the testicular degeneration and examined the mechanists) responsible for the gonadal toxicity and fertility effects. Singh et al. (1974) treated groups of young (8 to 10 weeks old) 25- to 30-g male mice with a single intraperitoneal injection of undiluted DEHP representing one-third, one-half, and two- thirds of the acute LDso (12,800, 19,200, and 25,600 mg/kg, respectively), and they subsequently bred these males to untreated females over a 12- week period. There was a reduction in the number of implantations, in litter size, and in the incidence of pregnancy at the high dose tested, particularly in the first few weeks after injection. Gray et al. (1977) reported significant decreases in the testicular weight of rats fed a 2.0% DEHP diet for 6 or 17 weeks and histological changes that included severe seminiferous tubular atrophy and cessation of sper-

Toxicity of Selected Contaminants 353 matogenesis. Lower dosage groups (0.2% and 1.0%) also exhibited hi~- tological changes in the testes, the incidence and severity of which correlated with the dietary concentration of DEHP. Testicular weights from rats that received a 0.2% diet were not reduced, but they did show histological evidence of reduced spermatogenesis with a reduction in late-stage sper- matids and mature sperm. The interstitial cells of the testes appeared undamaged histologically, but the authors described the appearance of "castration cells" (enlarged basophilic cells) in the pars distalis of the pituitary gland. The presence of castration cells in the pituitary is consid- ered to be a sensitive indication of gonadal deficiency (Russfield, 1967) and suggests that the testosterone-producing Leydig's cells of the testicular interstitium may have been damaged by exposure to dietary DEHP. In a later study, Gray and Butterworth (1980) reported that other phthal- ate esters produced testicular injury similar to that produced by DEHP: loss of the advanced germinal cells with only spermatogonia, Sertoli's cells, and occasional primary spermatocytes remaining. In this study, in which 4- and 10-week-old male rats received DEHP doses of 2,800 ma/ kg/day by intubation for 10 days, the testicular atrophy was not prevented by concurrent treatment with testosterone (200 ~g/kg/day) or follicle- stimulating hormone (FSH; 100 units subcutaneously), suggesting that DEHP does not interfere directly with hormone production or release. The testicular effects in 4-week-old rats that received dietary DEHP (2%, or approximately 1,200 mg/kg/day) were reversible, whether the treatment was stopped before or after the rats normally reached sexual maturity. The authors suggested that the histological appearance of the lesion and the kinetics of its reversibility make it unlikely that DEHP acted by direct cytotoxic action on the germ cells. Agarwal et al. (1985) fed male rats DEHP in dietary concentrations of 0, 320, 1,250, 5,000, or 20,000 ppm for 60 consecutive days. They reported a dose-dependent decrease in total body weight, testis weight, epididymal weight, and prostate weight at 5,000 and 20,000 ppm. In addition, degenerative changes in the testis, decreased epididymal sperm number and motility, and increased frequency of abnormal sperm forms were reported in the 20,000-ppm dose group. Reductions in fertility pa- rameters were correlated with gonadal effects, but they were not as severe. (Only average litter size was reduced in the 20, 000-ppm group. No re- duction in pup weights or growth rate and no increase in abnormalities were observed.) There were no fertility effects at lower doses. These data suggest that DEHP may also affect epididymal function, but not at dosages below those that produce minor seminiferous tubule degeneration. In a recent study, male rats received repeated intravenous infusions of a DEHP emulsion in doses of (5, 50, or 500 mg/kg) during a 3-hour period every other day for 12 days (Sjoberg et al., 1984~. The route of

354 DRINKING WATER AND HEALTH administration and dose concentrations (at the low dose, at least) used in this study are relevant to situations in which humans are exposed (i.e., generally by tranfusion with DEHP-contaminated blood or plasma). At the high dose, some degeneration of primary spermatocytes occurred, and ultrastructural changes were observed in the nuclei of late-stage sperrna- tids. No effects were found in the two lower dose groups. The author concluded that DEHP-infused intravenous solutions produced only very minor changes when low doses were used (and these were in the liver, not the testes), whereas the high doses resulted in distinct but not very severe changes. Since the germinal cells affected by DEHP (postmitotic spermatocytes and spermatids) are inside the Sertoli's cell barrier (Steinberger and Stein- berger, 1977), Gray and Butterworth (1980) proposed that phthalate esters may act as a result of injury to the Sertoli's cells. This suggestion is supported by the work of Foster et al. (1982), who described the histo- logical changes in the rat testes after oral administration of a single dose of di-n-pentyl phthalate (DPP), which produces testicular lesions in young male rats similar to those produced by DEHP. These changes, which occurred within 6 hours after dosing, included a severe vacuolation of Sertoli's cell cytoplasm and extended to degradative changes in the mi- tochondria of Sertoli's cells, spermatocytes, and spermatids within 24 hours. Recently, Gray and Beamand (1984) tested the effects of several phthal- ate esters, including DEHP, on primary cultures of testicular cells. Cultures consisted of a monolayer of Sertoli's cells to which clusters of sper- matogonia and spermatocytes were attached. The addition of MEHP to the culture medium resulted in a concentration-dependent increase in germ cell detachment over the range of 10-7 to 10-4 M. No effect was produced by the addition of DEHP or 2-EM. These observations suggested that the germ cell cultures lack the capability to metabolize DEHP to MEHP and 2-EM, and that 2-EH itself is without toxic effect. Testing with other phthalate monoesters revealed that only those esters that produce testicular injury in viva caused detachment of germ cells in culture. The authors suggested that the effects produced by the phthalate monoesters in culture may reflect damage to the Sertoli's cells. Gray and Gangolli (1986) reported that a single oral dose of DPP (2,200 mg/kg) or MEHP (1,000 mg/kg) administered in corn oil to 4- to 5-week- old male Sprague-Dawley rats markedly inhibited both the secretion of seminiferous tubular fluid and androgen-binding protein indicative of com- promised Sertoli's cell function. As in the earlier study, phthalate esters that did not cause testicular injury also did not affect Sertoli's cell function In VIVO.

Toxicity of Selected Contaminants 355 The sensitivity of the testis to the toxic effects of DEHP appears to be age dependent. Gray and Butterworth (1980) observed that treatment with DEHP (2% of the diet, or approximately 1,200 mg/kg/day) produced testicular lesions in 4- and 10-week-old rats but not in mature 15-week- old male rats, demonstrating that DEHP interferes with the normal de- velopment of rat testes but not with the mature testis. Curto and Thomas (1982) reported that mature male mice (35 to 40 g) were resistant to the effects of DEHP or its principal metabolite MEHP. Single intraperitoneal or subcutaneous doses of MEHP or DEHP up to 100 mg/kg failed to induce alterations in testicular or accessory sex gland weight in these animals. In a recent study (Seth and Mushtaq, 1984), DEHP was admin- istered orally (2,000 mg/kg) to male rats ranging in age from 4 to 12 weeks. Rats up to 8 weeks old had a 60% to 70% reduction in testis weight as well as changes in several biochemical parameters indicative of testicular damage. Rats older than 8 weeks exhibited no testicular weight reduction and no changes in any biochemical parameter, suggesting that developing males are more susceptible to the damaging effects of DEHP -r__ ~ than are mature males. Several investigators have studied the possible role of testicular zinc concentrations in mediating the effects of the phthalate esters on the testes. In rats, dietary zinc deficiency is known to cause testicular atrophy which appears to be due to inhibited pituitary testosterone output (Miller et al., 19601. In rats, phthalate esters causing testicular damage enhance urinary zinc excretion and lower testicular zinc concentrations (Foster et al., 1980, 19821. A 1-week dietary administration of 2% DEHP (Oishi and Hiraga, 1980a) or 2% MEHP (Oishi and Hiraga, 1980b) lowered testicular zinc concentrations in the rat. Foster et al. (1982) demonstrated that the loss of zinc following oral administration of DPP (2,200 mg/kg bw) occurred in the Golgi region of spermatids and was seen before there was any evidence of morphological damage in these cells. Curto and Thomas (1982) failed to detect any depletion of zinc in the testes of mature mice after intraperitoneal or subcutaneous treatment with up to 100-mg/kg doses of MEHP or DEHP, but they did see a reduction in gonadal zinc in the testes of mature rats after intraperitoneal administration of 100-mg/kg doses of DEHP. Cater et al. (1977) reported that coadministration of zinc by subcuta- neous injection substantially protected against the testicular damage pro- duced by a DPP dose of 2,000 mg/kg in rats, whereas Oishi and Hiraga (1983) did not observe any measure of protection from DEHP-induced damage (2,000-mg/kg doses of DEHP administered orally) when zinc was administered to male rats by diet or intraperitoneally. However, although liver and serum zinc concentrations were elevated in this study, testicular

356 DRINKING WATER AND H"LTH zinc concentrations were unchanged even in the control animals, leading the authors to conclude that the toxic effect of DEHP on the testis does not result from interference with gastrointestinal absorption of zinc. In subsequent studies, young rats fed dietary DEHP (2% for 1 week) de- veloped changes in testicular cholesterol and free fatty acid content as well as changes in fatty acid composition in phospholipids and triglycerides that were similar to animals fed a zinc-deficient diet (Oishi, 1984a). This testicular injury was not attributed to the vitamin A deficiency associated with lowered zinc concentrations in the testes (Oishi, 1984b). T. J. B. Gray et al. (1982b) and Gangolli (1982) reported the species specificity of testicular toxicity of the phthalate esters in young (4- to 6- week-old) males. Rats and guinea pigs were severely affected by oral administration of 2,000-mg/kg bw doses of DEHP for 10 days, mice were much less affected, and hamsters showed no effect. Curto and Thomas (1982) described similar differences in the sensitivity of mature mouse and rat testes to the effects of DEHP or MEHP treatment. Little attention has been paid to the gonadal toxicity of the phthalate esters in the nonpregnant female. Gray et al. (1977) reported that absolute ovary weights in female rats were decreased by a 17-week exposure to 2.0% dietary DEHP, but that there were no histological abnormalities in either the ovary or the pituitary. The authors suggested that DEHP may produce gonadal damage only in the male and that the decreased ovary weights were primarily a reflection of the reduced body weight of the treated females. Lower dosages (0.2% and 1.0%) had no effect. In an 18- week dietary exposure to 0.3% DEHP, Reel et al. (1985) found significant decreases in the reproductive tract (uterus and ovary) weight of female mice and complete suppression of fertility in these females when bred to unexposed males immediately following the 18-week exposure. These data demonstrate that DEHP is toxic to the reproductive tract of female and male rodents, but only at very high doses. CONCLUSIONS AND RECOMMENDATIONS The committee believes that any chemically induced, measurable al- teration in normal biochemistry or physiology should be viewed with concern, even if the resulting health effects are not immediately appre- ciated. Because DEHP has known effects on the liver, a risk assessment is presented for alterations in liver function. As noted earlier, Ganning et al. (1984) found that dialysis patients who were receiving approximately 150 mg of DEHP intravenously each week as a consequence of their treatment experienced no reported morphological changes in the liver, but at one year, peroxisomes were "significantly

Toxicity of Selected Contaminants 357 TABLE 9-17 Increases in Liver Function in Male Sprague-Dawley Rats after Exposure to DEHPa Dietary Concentrations (ppm) Change 50 100 500 1,000 2,500 5,000 Liver wet weight Ncb NC NC +c + + Peroxisomes NO NR NC NR NR + Hepatic catalase NR NC NR NC NR + Hepatic can~itine acetyltransferase (CAT) NR NC + + + + Hepatic carnitine palmitoyltransferase (CPI) NR NC + + + + Hepatic 2-oxidation capacity NC NC + + NR + Hepatic ~glycendes NC NC NC NC NC NC aAdapted from Morton, 1979. bNC = no change compared with controls. c + = statistically significant, p ' 0.05. OR = not reported. higher in number." For a 70-kg person, the exposure would be approx- imately 2. 1 mg/kg/week, or 0.3 mg/kg/day. This study population includes persons who are not in peak condition and who have an undue burden of toxicants. If a 10-fold safety factor were applied to the dose level in addition to a 10-fold safety factor for person-to-person variability, a value of 0.003 mg/kg/day would be obtained. In a study by Morton (1979), male Sprague-Dawley rats were fed diets containing 0- to 5,000-ppm concentrations of DEHP for 7 days. The resultant increases in liver enzymes are listed in Table 9-17. When all indices of altered liver function are considered as a whole, the no-observed-effect level (NOEL) in this series of studies appears to be approximately 50 ppm (3.30 mg/kg bw a day) in the diet. If a safety factor of 1,000 is applied because this was only a short-term animal study, the acceptable daily intake would be 0.003 mg/kg. In comparison, Ganning et al. (1984) fed rats DEHP for 2 years in the diet and found alterations in palmitoyl-coenzyme A dehydrogenase and dolichol in the liver. The lowest-observed-effect level (LOEL) for this study was 0.02% (200 ppm or 13 mg/kglday). If a safety factor of 100 is applied in addition to a safety factor of 10 because a LOEL is being used, the acceptable daily intake would be 0.01 mg/kg. The committee noted that the areas of concern for possible toxic effects induced by DEHP are changes in liver function, reproductive and fertility effects, developmental effects, and cancer. In accordance with the guide-

358 DRINKING WATER AND HEALTH TABLE 9-18 Summary of Estimated Acceptable Daily Intakes (ADIs) for DEHP Estimated ADIsa Type of Study (mg/kg/day) Reference Human, dialysis patients 0.003 Ganning et al., 1984 Rat, liver function 0.003 Morton, 1979 Rat, liver enzymes 0.01 Ganning et al., 1984 Rat, teratogenicity study 0.50 Ruddick et al., 1981 aThe committee noted that in the NTP report (1983b) a 0.032-mg/kg/day dose produced a 1 x 10-6 risk of cancer in a bioassay for carcinogenesis in mice. lines in Chapter 8, it suggests that the LOEL or NOEL and a judgmental safety factor approach should be used for the first three end points. A linearized multistage model should be applied in cancer risk assessment. The estimated acceptable daily intakes shown in Table 9-18 were derived by the different procedures and from the different studies noted above. The most sensitive responses from the estimates appear to be liver per- oxisomal proliferation, function, and tumor production. The teratologic effects are the least sensitive of the responses examined. The mechanism of action of DEHP carcinogenicity is not known, and the contribution of genetic toxicity to its carcinogenicity is unclear. The only genotoxic activities observed in cell culture models were the induction of aneuploidy, Syrian hamster embryo-cell transformation, and some small but significant increase in SCEs. Short-term in viva tests for genetic tox- icity have thus far been negative. Activities alone or in combination that may contribute to carcinogenicity include promotional activity, forced cell proliferation, and genetic effects resulting from peroxisomal proliferation. Several properties of DEHP, e.g., induction of peroxisomal proliferation, merit further investigation to determine whether there should be concern about this and related compounds. Experiments that show lack of genotoxic activity in peroxisomally proliferated animals and examples of noncarcino- genic peroxisome proliferators (DEHA in the rat) indicate that peroxisomal proliferation is not a sufficient event for complete carcinogenesis. Nonethe- less, since a number of hypolipidemic agents induce peroxisomal proliferation in rodents, further work should be done to evaluate the role of this process in carcinogenesis and its risk to humans. Differences between species in the rates of peroxisomal proliferation and its extent have been observed (Rhodes et al., 1986; Reddy and Lalwani, 1983~. Microbodies (peroxisomes) were observed in humans following long- term administration of the hypolipidemic agent clofibrate, but the effect was less pronounced than that observed in rodents (Ready and Lalwani, 19831. Kidney patients receiving DEHP in the form of a contaminant

Toxicity of Selected Contaminants 359 during the process of dialysis had an increase in hepatic peroxisomal proliferation after a year (Ganning et al., 19841. Results of studies with primary hepatocyte cultures show that DEHP and the metabolites respon- sible for peroxisomal proliferation induce peroxisomal enzymes in rat hepatocytes but not in human hepatocytes (T. J. B. Gray et al., 1982a; A. M. Mitchell et al., 1984; Rhodes et al., 19861. The ability of hypo- lipidemic agents to induce dramatic changes in hepatocytes, including cell proliferation, indicates receptor-mediated regulation of many genes (Ready and Lalwani, 19831. This could possibly play a role in carcinogenesis. The extent to which this occurs in human cells should be examined further. The ability of DEHP to transform Syrian hamster embryo cells and induce aneuploidy and possibly SCEs is reminiscent of other human car- cinogens that do not show mutagenic activity. Surfactants as a whole appear to be without genotoxic activity in cell culture assays (Yam et al., 19841. The structure-activity relationships of DEHP, MEHP, and similar surfactants with membrane-altering properties should be examined in cell transformation bioassays and correlated with the ability to induce aneu- ploidy and, if possible, carcinogenic activity. ETHYLENE DIBROMIDE I,2-Dibromoethane CAS No. 106-93-4 RTECS No. KH9275000 H H 1 1 H—C—C—H 1 1 Br Br Since ethylene dibromide (EDB) was reviewed in Volume 3 of Drinking Water and Health (NRC, 1980, pp. 98-101), the following discussion is mainly an examination of data that have since become available to the committee. HEALTH ASPECTS Observations in Humans Information concerning the health effects in humans resulting from exposure to EDB are limited to descriptions of acute symptoms observed after episodes of relatively high-dose exposure (NIOSH, 1977a); an epi- demiological study of limited size regarding cancer incidence among work- ers engaged in EDB production (Ott et al., 19801; and several evaluations of reproductive performance of such workers (Ter Haar, 1980; Wong et

360 DRINKING WATER AND H"LTH al., 19791. Acute dermal exposure produces painful local inflammation, swelling, and blistering. Acute systemic exposure by respiratory or oral routes causes vomiting, diarrhea, abdominal pain, and in some cases delayed lung damage and depression of the central nervous system. In one instance, a woman who ingested 4.5 ml of EDB died from massive centrilobular necrosis of the liver and focal damage to renal proximal tubular epithelium (Olmstead, 19601. In addition, two fatalities were re- ported after an acute occupational exposure of workers cleaning a storage tank. Both workers had metabolic acidosis, hepatic damage, and acute renal failure (Letz et al., 19841. In an epidemiological study, Ott et al. (1980) examined the mortality experience of 161 white males employed in various phases of EDB syn- thesis at two production plants from 1940 through 1976. Numbers were limited, and the reported excess mortality from cancer was not statistically significant (7 cases observed, 5.8 expected). Average levels of workplace exposure to EDB were estimated to have ranged from 0.4 ppm in air for these two cohorts of workers. The power of this study to find a risk of twofold or more is roughly 63C%o, which is below the usually minimum power of 90%. In a study of reproductive performance, Wong et al. (1979) evaluated the reproductive experience, from 1958 to 1977, of 297 EDB-exposed men employed in four EDB manufacturing facilities and their wives. Subjects were divided into two exposure groups: less than 0.5 ppm and between 0.5 and 5.0 ppm. The number of live births was determined for each group and compared to the national average. The authors concluded that overall, the standardized birth ratio was not significantly reduced among EDB-exposed workers and their wives, and no evidence of a dose- response relationship was seen. Ter Haar (1980) reviewed a number of similar studies, also concluding that there was no evidence of EDB-related reproductive hazard. Observations in Other Species Acute Effects There is information on the acute lethality of EDB in a number of species. Oral LDso values reported by Rowe et al. (1952) range from 55 mg/kg bw in female rabbits to 420 mg/kg bw in female mice. The LDso values of Rowe et al. (1952) for male and female rats are 146 and 117 mg/kg, respectively. Although animal studies have confirmed that EDB is acutely toxic to the liver and kidneys, near-lethal doses are apparently required to cause observable organ damage. Storer and Conolly (1983) saw little evidence of liver or kidney damage in male B6C3F~ mice given a single intraper-

Toxicity of Selected Contaminants 361 itoneal injection of 94 or 141 mg/kg. An intraper~toneal dose of 188 ma/ kg was required to produce significant increases in liver and kidney weight as well as increases in blood urea nitrogen and serum enzyme levels. Nachtomi and Alumot (1972) observed an increase in total liver lipids and triglycerides in fasted male rats given EDB doses of 1 10 mg/kg bw orally and a modest increase in lipid peroxidation of microsomal liver lipids. In a subsequent study, the 110-mg/kg dose was found to produce central and midzonal hepatic necrosis in fasted male rats (Broda et al., 19761. Acute oral no-observed-effect levels (NOELs) and lowest-observed-effect levels (LOELs) for injury of the liver, kidneys, or any other organ have not been identified. There have been a number of investigations of biochemical changes in the livers of EDB-dosed rats. Moody et al. (1981) found that a large oral dose (220 mg/kg bw) of EDB produced a drop in hepatic microsomal cytochrome P450 levels but no significant change in fatty acids. Moody et al. (1982a) extended their findings by demonstrating significant de- creases in the cytochrome P450 content of microsomes isolated from the liver, kidney, testes, and lungs of fasted male rats given a single oral dose of 159 mg/kg. Little or no effect was observed for other microsomal parameters, including dealkylations catalyzed by mixed-function oxidases. Botti et al. (1982) administered 75-mg/kg bw doses of EDB orally to fasted male rats and examined the time course of changes in glutathione levels. Liver glutathione levels diminished within-15 minutes and remained low for up to 4 hours after dosing. Cytosolic glutathione-S-transferase activities did not diminish until 2 hours after dosing and returned to normal by 4 hours. The glutathione levels decreased due to metabolism of EDB by glutathione-S-transferases. When glutathione levels fell too low, re- active EDB-glutathione conjugates and other EDB metabolites were ap- parently free to interact with (and inhibit) the transferase enzymes as well as other proteins, RNA, and DNA. Nachtomi and Sarma (1977) found that 4 hours after oral dosing of rats with i4C-labeled EDB, the radioactivity was incorporated in decreasing amounts as follows: cytoplasmic protein > nuclear protein > microsomal protein—RNA > nuclear DNA. The peak occurrence of single-strand DNA breaks coincided with a substantial decrease in hepatic glutathione levels. Kowalski et al. (1985), using single, varying intravenous or in- traperitoneal doses of EDB with rats and mice, recently observed that several tissues, mainly the target tissues for the EDB-induced carcinogenic effects, can metabolize EDB to reactive products that become irreversibly bound to tissue constituents. White et al. (1981) reported a dose-dependent increase in single-strand DNA breaks in hepatic cells of mice given 25-, 50-, or 75-mg/kg doses of EDB by intraperitoneal injection. The extent

362 DRINKING WATER AND HEALTH of DNA damage was statistically significant at the two highest dosage levels. White et al. ~ 1981 ~ found no evidence of EDB-induced DNA cross- links or DNA-protein cross-links. Subacute Elects Few subacute toxicity studies have been conducted with oral administration of EDB. In one study, EDB was given in corn oil by Savage 5 times weekly for 6 weeks to male and female mice and rats in order to establish the maximally tolerated doses of EDB for use in a subsequent carcinogenesis bioassay (NCI, 1978b). A dosage level of 100 mg/kg/day caused a reduction in body weight gain and 20% mortality in male and female rats. Male mice receiving 63 and 159 mg/kg/day exhibited a dose-dependent reduction in body weight gain. Nachtomi (1980) administered fumigated feed containing 100 and 200 ppm EDB to male rats for 18 days. Consumption of these diets was said to be equivalent to a daily intake of 10 to 20 mg/kg. The dietary regimen produced no change in liver weight or DNA concentration, compared to controls, but DNA synthesis was slightly increased. This increase was not dose dependent, and it was not large enough to be statistically significant. The subacute toxic potential of inhaled EDB has been investigated in several species. Rowe et al. (1952) subjected male and female rats, rabbits, guinea pigs, and monkeys to as many as 156 7-hour exposures to 25-ppm concentrations of EDB for periods as long as 220 days. There was no evidence of adverse effects in any species. Repeated 7-hour exposures to 50 ppm for up to 91 days, however, were not well tolerated. Guinea pigs were the most sensitive, in that they had depressed body weight gain; increased lung, liver, and kidney weights; slight central fatty degeneration of the liver; and slight degeneration of the renal tubular epithelium. Reznik et al. (1980b) exposed male and female Fischer 344 rats and B6C3F~ mice to 0-, 3-, 15-, or 75-ppm concentrations of EDB vapor 6 hours/day, 5 days/week for 13 weeks. Histopathological changes were limited to the respiratory tract in animals sacrificed at the end of 13 weeks. Most of the male and female mice and rats subjected to 75-ppm concen- trations of EDB exhibited nasal cavity changes, including loss of cilia, cytomegaly, focal hyperplasia, and squamous metaplasia. No mice, and only one or two rats, developed these changes in the 15-ppm groups. Morphological alterations were not seen in the control or the 3-ppm groups. Nitschke et al. (1981) also conducted a 13-week inhalation study of EDB. Male and female Fischer 344 rats were exposed to 0-, 3-, 10-, or 40-ppm concentrations of EDB 6 hours/day, 5 days/week for 13 weeks. Groups of animals were sacrificed after 1, 6, and 13 weeks of exposure as well as after a recovery period of 88 to 89 days. Male rats exposed to 40-ppm concentrations of EDB experienced decreased body weight gain and increased liver and kidney weight. There were no changes in clinical

Toxicity of Selected Contaminants 363 chemistry parameters and only slight, transient alterations in some he- matology and urinalysis indices in female rats. The most sensitive index of EDB inhalation exposure was a morphologic change of the nasal respiratory epithelium, which was observed as early as 1 week in the 10- ppm-dosed animals. The changes progressed in the 40-ppm group from scattered to diffuse hyperplasia at 1 week to diffuse or focal nonkeratin- izing squamous metaplasia and hyperplasia with focal cell necrosis at 13 weeks. There was a reversion to normal within the 88- to 89-day recovery period in both the 10- and 40-ppm groups. Chronic Elects The chronic toxicity of EDB has been assessed in conjunction with two carcinogenicity studies. NCI (1978b) reported the results of a cancer bioassay in which male and female Osborne-Mendel rats and B6C3F~ mice were given EDB in corn oil by gavage 5 times weekly. Although the study was originally scheduled to last approximately 2 years, it was terminated early due to excessive death rates. The doses were reported as time-weighted averages (TWAs), since both dosage levels and dosage schedules were changed during the course of the investigations. The TWA for the low- and high-dose male and female rats was approx- imately 40 mg/kg. The TWA for the low-dose male and female mice was 62 mg/kg; for the high-dose mice, it was 107 mg/kg. The low- and high- dose mice and rats had decreased body weight gain and decreased survival, relative to controls. Approximately 40% of the high-dose male and female rats died in week 15, ostensibly due to "acute toxic reactions." Eleven of the 18 high-dose males dying at the time exhibited acanthosis and hyperkeratosis of the forestomach. Pathological changes observed in other organs of low- and high-dose male and female rats included degenerative changes in the liver and adrenal cortex. Both male mice and rats exhibited testicular atrophy. Unfortunately, the excessive mortality and alterations in the dosage regimen preclude identifying a NOEL or a LOEL in the NCI (1978b) study. In the second chronic-effects study, Wong et al. (1982) exposed male and female Sprague-Dawley rats to 20-ppm concentrations of EDB vapor 7 hours/day, 5 days/week for 18 months. Body weights of five EDB- exposed rats were slightly lower than those of the controls throughout the study period, but the decreases were not always sufficient to be statistically significant. Increased mortality in both males and females inhaling EDB was manifest at 15 and 18 months. Hematological studies performed on the moribund animals yielded normal results. No mention was made of histopathological evidence of toxicity, other than the finding of a higher tumor incidence in EDB-exposed rats than in controls. Apparently, the increased tumor incidence was responsible for the increased mortality.

364 DRINKING WATER AND HEALTH EDB has been found to interact synergistically with disulfiram (Anta- buse) to produce greatly increased toxicity (NIOSH, 1978b). In laboratory rats exposed simultaneously to both EDB by inhalation (20 ppm) and disulfiram in the diet (0.05% by weight in the feed), there was excep- tionally high mortality after approximately 13 months of exposure when compared to animals exposed to either EDB or disulfiram alone. Fur- thermore, in a chronic carcinogenicity study (NIOSH, 1981) of rats ex- posed to both substances, EDB by inhalation and disulfiram in the feed, there was an approximately 10-fold increase in the incidence of hepato- cellular carcinomas compared to the rate in animals exposed to either substance alone. Mutagenicity Moriya et al. (1983) monitored the dose response of EDB-induced mutation in the Salmonella typhimurium bacterial reverse mutation assay by using several tester strains as well as Escherichia cold WP2 her. They found that EDB is mutagenic at a very high dose and that the mutation rate is not enhanced by the addition of rat liver microsomal extract. Unfortunately, the results were expressed as the number of re- vertants per plate, not revertants per survivors, so toxicity was not as- sessed. The authors concluded that EDB can be a direct-acting mutagen in bacteria. That conclusion is consistent with an earlier, less detailed study by Buselmaier et al. (1972), which was reviewed in Volume 3 of Drinking Water and Health (NRC, 1980, p. 991. Ohta et al. (1984) dem- onstrated that EDB induces the SOS (error-prone DNA repair) response in E. cold tester strains and that this response does not require microsomal activation. The authors concluded from those data that EDB behaves as a direct-acting mutagen with respect to the error-prone DNA repair re- sponse. In contrast, van Bladeren et al. (1980) found that the addition of the 100,000 x g microsomal supernatant fraction considerably enhanced the mutagenicity of EDB in the S. typhimurium TA100 strain. The authors concluded that the primary glutathione adduct is responsible for the mu- tagenic effect. When treated with gaseous EDB, Drosophila develop sex-linked lethal mutations (Kale and Baum, 1983) that appear at acute and chronic ex- posure levels as low as 2.3 ppm/hr. The authors concluded that EDB can be mutagenic in a eucaryote and that such a mutation may be a general property of the EDB interaction. Although some bacterial assay systems suggest that EDB can be a direct- acting mutagen, EDB may be activated chemically or enzymatically in eucaryotic cells. DiRenzo et al. (1982) showed that in an in vitro assay EDB is activated by hepatic microsomal extracts to form a species that can bind covalently to DNA. Similarly, van Bladeren et al. (1981) showed

Toxicity of Selected Contaminants 365 that EDB may be conjugated with glutathione in microsomes to form thiiranium ions, which might then bind covalently to DNA or other target macromolecules . Carcinogenicity The NCI bioassay ~ 1978b) of EDB showed that, when administered intragastrically, EDB produces squamous cell carcinomas of the forestomach in Osborne-Mendel rats and, to a lesser extent, in B6C3F~ mice. For this experiment, EDB doses began at 80 and 40 mg/kg/day for rats, but after 16 weeks the higher dose level was discontinued for 14 weeks due to toxicity. At week 30, the high-dose group was placed on a 40-mg/kg/day regimen. Administration of EDB to the lower-dose group was continued without interruption. For mice, the original dose levels were 120 and 60 mg/kg/day. After 13 weeks, however, a 2-week trial of 200 and 100 mg/kg/day was begun. At week 15, the original dose levels were given, but after 42 weeks all mice were changed to one level of 60 mg/kg/day. Doses were ultimately reported as time-weighted averages of 0, 37, or 39 mg/kg/day for female rats; 0, 38, or 41 mg/kg/day for male rats; and 0, 62, or 107 mg/kg/day for male and female mice. The tumor rates for male rats were 33/50 and 45/50, whereas the rates for female rats were 29/50 and 40/50 for the high- and low-dose groups, respectively. Tumors invaded the forestomach locally, eventually metas- tasizing throughout the abdominal cavity. The authors believed that be- cause so many animals died early due to the extensive toxicity of the compound, the incidence of tumor development was reduced. No fore- stomach tumors were observed in control groups. An increased incidence of liver cancer and hepatic nodules was also observed in rats, especially among the high-dose females. Hemangiosarcoma was observed among male rats. In mice, squamous cell adenocarcinomas of the forestomach occurred in 29 of the 49 high-dose males, 28 of the 50 high-dose females, 45 of the 50 low-dose males, and 46 of the 49 low-dose females. None was found in control groups. Lung cancers were observed in 10 of the 47 high- dose males, 6 of the 46 high-dose females, 4 of the 45 low-dose males, and 10 of the 43 low-dose females. No lung tumors were seen in control mice. In another carcinogenesis bioassay (NTP, 1982b), Fischer 344 rats and B6C3F~ mice were chronically exposed to 10 or 40 ppm EDB by inhalation for 6 hours/day, 5 days/week. Increased mortality was observed for the high-dose rats and for mice at both doses. Male rats developed tumors of the nasal cavity with the following incidences: 20 adenocarcinomas and 3 squamous cell carcinomas in the 50 low-dose animals, and 28-adeno- carcinomas, 21 carcinomas, and 3 squamous cell carcinomas in the 50 high-dose animals. In the nasal cavity of female rats, there were 20 ad-

366 DRINKING WATER AND HEALTH enocarcinomas and 1 squamous cell carcinoma in the 50-animal low-dose group, and 29 adenocarcinomas, 25 carcinomas, and 5 squamous cell carcinomas in the 50-animal high-dose group. One female control rat developed a squamous cell carcinoma of the nasal cavity. Hemangiosar- comas were also associated with EDB exposure, occurring in 1 of 50 low- dose males, 15 of 50 high-dose males, and 5 of 50 high-dose females. The incidence of testicular mesotheliomas among males and mammary fi'oroadenomas among females was also increased by EDB exposure. In mice exposed to EDB by inhalation, only 6 of 50 high-dose females exhibited an incidence of nasal cavity carcinoma. However, alveolar/ bronchiolar carcinomas were found in 3 of 48 low-dose males, 5 of 49 low-dose females, 19 of 46 high-dose males, and 37 of 50 high-dose females. One control female developed this type of tumor. Occurrences of hemangiosarcomas, fibrosarcomas, and malignant mammary neoplasms were also associated with EDB exposure. Van Duuren et al. (1979) showed that EDB applied three times weekly to the dorsal skin of male and female Swiss mice at doses of 25 or 50 mg per application yielded significant incidences of skin papillomas, skin carcinomas, and lung tumors. In the 30-animal low-dose group, 2 had skin papillomas, 2 had skin carcinomas, and 24 had benign papillomas of the lung. In the 30-animal high-dose group, 8 had skin papillomas, 3 had skin carcinomas, and 26 had lung papillomas. These investigators also found that a single 75-mg dermal application of EDB followed by repeated applications of the promoter phorbol myristate acetate did not cause tumors. Wong et al. (1982) conducted an inhalation study of EDB exposure with Sprague-Dawley rats. (This study was described above in the chronic effects section.) As compared to controls, there were significant (p < 0.05) incidences of spleen hemangiosarcomas (10 of 46 males, 6 of 48 females), adrenal tumors (11 of 46 males, 6 of 48 females), subcutaneous mesenchymal tumors (11 of 46 males), and mammary tumors (25 of 48 females). A carcinogenic risk estimate was not attempted tor ~~u tor several reasons. The results of the carcinogenicity study in which EDB was ad- ministered by gavage to rats and shown to cause squamous cell carcinomas of the forestomach are quantitatively equivocal. Following the extensive toxicity caused by the higher dose of EDB early in the experiment, treat- ment was discontinued for 14 weeks and then reinstituted at a lower dose. The total dose is thus expressed as a time-weighted average and as such is not representative of a situation in which humans are chronically exposed to a low dose of the chemical. When expressed as time-we~ighted averages, the doses given to the low- and high-dose groups are virtually identical. .. . . ~ _~ ~ ~

Toxicity of Selected Contaminants 367 The inversion in tumor yield can be accounted for only by early toxicity and does not represent a reliable dose response. The inhalation study also had some of the same shortcomings as the Savage study, such as increased early mortality. In addition, inhalation was considered less relevant to the estimation of risk for drinking water than the oral route. Reproductive Eff;ects Early studies in hens demonstrated that EDB causes decreased egg size, impaired uptake of labeled proteins by ovarian follicles, egg infertility, and cessation of egg laying (Alumot and Harduf, 1971; Alumot et al., 1968; Bondi et al., 1955~. Amir et al. (1977) reported that oral administration of 10 EDB doses of 4 mg/kg bw produced a large but transient decrease of sperm count and a large but transient increase in the number of abnormal sperm in bulls. More recently, in- the NTP (1982b) inhalation bioassay described in the carcinogenicity section, tes- ticular degeneration was found in 10 of 50 low-dose and 18 of 49 high- dose rats, and testicular atrophy was observed in 2 of 50 low-dose and 5 of 49 high-dose rats. (EDB was administered at doses of 10 or 40 ppm for 6 hours/day, 5 days/week.) Among the 50 control rats, testicular degeneration occurred in one animal and atrophy was observed in another. Testicular changes were not reported for EDB-exposed mice (NTP, 1982b). In the NCI carcinogenesis bioassay of EDB by Savage, early devel- opment of testicular atrophy was observed in dosed rats and high-dose mice (NCI, 1978b). Administration of three doses of EDB by inhalation (Short et al., 1979) to both male and female rats for 10 weeks resulted in adverse effects on reproduction in both sexes exposed to the highest dose (89 ppm for males and 80 ppm for females). Mortality and morbidity were observed as well, however, so it was not possible to attribute re- productive effects directly to EDB. When EDB was administered by in- halation (20 ppm) in the presence of dietary disulfiram (0.05%), Wong et al. (1982) found that 90% of Sprague-Dawley rats developed testicular atrophy. Edwards et al. (1970) found that intraperitoneal administration of five daily 10-mg/kg doses of EDB impaired spermatogenesis in Wistar rats. Motor reflexes and motor coordination were assessed in the F~ prog- eny of male rats treated intraperitoneally for 5 days with EDB doses of 1.25, 2.5, 5.0, or 10.0 mg/kg. Although litter size remained unchanged, several indices of behavior were affected. Doses of EDB that produced these behavioral effects were much less than those (close to the LD50) required to produce abnormal sperm morphology (Fanini et al., 19841. Together these data suggest that like haloalkanes, EDB appears to induce testicular damage at low doses in experimental animals. Although evidence suggests that EDB may also induce sperm death and sperm abnormalities, those data are incomplete.

368 DRINKING WATER AND HEALTH CONCLUSIONS AND RECOMMENDATIONS Ethylene dibromide is mutagenic, carcinogenic, and toxic to the repro- ductive system of laboratory animals. In some bacterial in vitro assay systems, it appears to be a direct-acting mutagen. Nonetheless, it can be activated by hepatic microsomal extracts to form glutathione adducts and bind with DNA. EDB produced squamous cell carcinomas- of the fore- stomach, which metastasized throughout the abdominal cavity in rats and mice. In addition, lung tumors were observed in mice gavaged with EDB. Inhalation studies with EDB produced nasal cavity tumors and tumors at various other sites. A carcinogenicity risk estimate was not calculated for EDB because of the lack of reliable dose-response data from the single oral exposure study available. N ITROFEN 2,4-DichIoro- ~ -~4-nitrophenoxy~benzene CAS No. IS36-75-5 RTECS NO. KN8400000 Cl C1 40 /3NO2 Nitrofen is a selective contact herbicide used on a variety of food crops to control annual grasses and weeds both before and after the crops begin to grow. The Rohm and Haas Company obtained the first registration for vegetable crop use in 1966, and the first European authorizations followed the next year. This compound is a member of the chlorophenoxy class of herbicides and is also called nitrophene, TOK, TOK E-25, and Nip (Burke Hurt et al., 1983~. Before 1981, nitrofen was registered in the United States for the control of annual bluegrass, crabgrass, goosefoot, iamb' s-quarters, malva, night- shade, nettle, pigweed, purslane, shepherd's purse, and spergularia on broccoli, Brussels sprouts, cabbage, cauliflower, carrots, celery, garlic, horseradish, onions, parsley, sugar beets, carnations, chrysanthemums, roses, stocks, ground covers, and noncropped lands. Its principal use has been as a herbicide to control malva or cheeseweed in California cole crops and purslane in the onion fields of New York and the upper Midwest (Burke Hurt et al., 19831. Food tolerance levels of nitrofen ranging from 0.02 to 0.75 ppm have been established for 24 commodities, including vegetables, milk, rice, poultry, and meats (EPA, 1984a). The estimated

Toxicity of Selected Contaminants 369 annual production was between 227,000 and 454,000 kg in 1974 (Kim- brough et al., 1974~. Nitrofen is used to control weeds in peas, beans, and spring wheat and other cereal grains in Northern Europe and for a variety of vegetables in the Mediterranean regions. In China, Japan, and the Far East, it is used principally on rice. Effective applications are generally between 2 and 6 kg/ha applied as a thin layer to the soil surface or to actively growing weeds at the 2- to 4-leaf stage (Burke Hurt et al., 1983~. Technical-grade nitrofen contains approximately 95% active ingredient and impurities consisting of approximately 3% p-chloronitrobenzene, 1% dichlorophenol, and 1% unknowns. Woolson et al. (1972) specifically determined that technical-grade nitrofen lacked 2,3,7,8-tetrachloro- dibenzo-p-dioxin contaminants. The technical-grade material is a dark brown solid with a slight aroma, a melting point of 64°C to 71°C, and a vapor pressure of 8 x 10-6 mm mercury at 40°C to 4 x 10-4 mm mercury at 70°C. It is relatively insoluble in water (0.7 to 1.2 mg/liter at 22°C) but readily soluble in most organic solvents (Ambrose et al., 19711. The odor threshold in water was determined to be 0.58 to 1.38 mg/liter, and was found to be highly stable (Shardina, 1972~. Nitrofen was detected at levels of less than 0.2 mg/m3 in samples of natural running water around areas of commercial use in Belgium (Deleu and Copin, 19791. METABOLISM Nitrofen spread on soil surfaces is adsorbed into weeds as they emerge. The compound is activated by sunlight and kills weeds by inhibiting photosynthesis. Activation for herbicidal activity appears to involve, but not be exclusively related to, cleavage at the ether linkage to form dichlo- rophenols and nitrophenol (Hawton and Stobbe, 19711. The biochemical mechanism of herbicidal activity is based upon interference with both oxidative and photosynthetic phosphorylation in mitochondria and chlo- roplasts of plants. Nitrofen and other diphenyl ether herbicides have been shown to inhibit chloroplast noncyclic electron transport by removing or inactivating an electron carrier associated with photosystem II of photo- synthesis, or the Hill reaction. Mitochondrial respiration appears to be disrupted at several sites along the electron transport chain (Moreland et al., 1970~. HEALTH ASPECTS Observations in Humans No data are available.

370 DRINKING WATER AND HEALTH Observations in Other Species Acute Elects The acute toxicity of nitrofen in laboratory animals has been summarized by Burke Hurt et al. (1983), who reported oral LDsoS from intragastric administration ranging from 2.4 to 3.6 g/kg in rats and mice. Lower values of 740 mg/kg for rats and 450 mg/kg for mice were cited by Ivanova (19671. The oral LDso by the dietary route has been reported to range from 410 to 3,580 mg/kg bw in rats (Ambrose et al., 1971; Kimbrough et al., 1974; Shardina, 1972; Wiswesser, 1976) and from 780 to 1,620 mg/kg in rabbits (Shardina, 1972; Wiswesser, 1976~. Burke Hurt et al. (1983) reported the dermal LDsoS for rats and rabbits to be 5 and 3.27 g/kg, respectively, whereas Ivanova (1967) found the percutaneous LDso for rabbits to be greater than 5 g/kg. A 1-hour inhalation exposure of rats to 271 mg/m3 decreased body weight gain but did not cause signs of toxicity. On the basis of inhalation studies with cats, Ivanova (1967) recommended that the maximum permissible concentration of nitrofen in the air of occupational settings should not exceed 1 mg/m3. Clinical manifestations of poisoning with acute exposure are characterized by decreased activity, a rise in the respiratory rate and disturbances in respiratory rhythm, and hyperthermia, proceeding to tremors and con- vulsions. Death from acute exposure generally occurs within 2 to 8 days (Burke Hurt et al., 1983; Ivanova, 19671. Subchronic Effects The subchronic toxicity of nitrofen in laboratory animals has been summarized by Burke Hurt et al. (19831. Wistar rats were fed diets containing 0-, 100-, 500-, 2,500-, 12,500-, or 50,000-ppm concentrations of nitrofen in a 13-week study by Ambrose et al. (19711. In the 50,000-ppm group, none of the rats survived the first week, and survival and body weight gain were affected in the 12,500- and 2,500- ppm groups. At 500 ppm (25 mg/kg) and lower, there were no treatment effects on growth, food consumpt on, or mortality. Dose-related increases in relative liver weights were observed at 100 ppm (5 mg/kg) and higher. In another study with dietary doses ranging from 0 to 2,500 ppm for 13 weeks, dose-related increases in liver, testes, and kidney weights occurred at 500 ppm (25 mg/kg) and higher (O'Hare et al., 19834. Elevated liver weights as well as induction of cytochrome P450 activity levels have been consistent, early signs of low-dose exposure of rodents (Burke Hurt et al., 1983~. Chronic E~ects The only chronic effect studies found were limited to . . . carc~nogen~c~ty .

Toxicity of Selected Contaminants 371 Carcinogenicity A bioassay of technical-grade nitrofen for possible carcinogenicity was conducted using Osborne-Mendel rats and B6C3F~ mice (NCI, 1978c). Nitrofen was administered in the feed at two different concentrations to groups of 50 male and 50 female animals of each species. Two of the 50 female rats fed 1,300 ppm developed a metastatic and locally invasive ductal carcinoma of the pancreas as did 7 of the 50 rats fed 2,600 ppm (which was statistically significant). This rare tumor was not observed in the simultaneous or pooled control groups from more than 200 bioassays (Milman et al., 19781. Poor survival precluded the evalu- ation of carcinogenicity in male rats (NCI, 1978c). In mice of both sexes, the incidence of hepatocellular carcinoma was statistically elevated at both treatment levels. This lesion occurred in 36 of 49 (73%) males in the 2,348-ppm group and in 46 of 48 (96%) males in the 4,696-ppm group, compared with 4 of 20 (20%) in simultaneous controls and 9 of 72 (12%) in pooled controls. In females, hepatocellular carcinomas occurred in 36 of 41 (88%) low-dose and 43 of 44 (98%) high-dose animals, whereas no such carcinomas were observed in simul- taneous (19) or pooled (80) controls. There was a significant positive association between dose and elevated incidence of hemangiosarcomas for male and female mice compared with pooled controls. The results of this study (NCI, 1978c) in Osborne-Mendel rats and B6C3F~ mice indicate that orally administered technical-grade nitrofen is carcinogenic in the livers of mice of both sexes and in the pancreas of female rats. Another bioassay to assess the carcinogenicity of nitrogen was conducted using Fischer 344 rats and B6C3F~ mice (NCI, 19791. The same design was used as in the previous study, except the high and low dietary con- centrations of nitrofen were 6,000 and 3,000 ppm for both species. A dose-related depression in body weight gain was observed in rats of both sexes, but the exposures had no effect on survival. Furthermore, there were no positive associations between treatment and tumor incidences in rats of either sex. In mice, dose-related decreases in body weight gain were observed for both sexes, but treatment had no effect on survival. Although a variety of tumors was noted, only the incidence of liver tumors was related to nitrofen exposure. Hepatocellular adenomas were seen in 1 of 20 (5%) control males, in 18 of 49 (37%) low-dose males, and in 20 of 48 (42%) high-dose males; hepatocellular carcinomas were found in none of the 20 controls, in 13 of 49 (26%) low-dose males, and in 20 of 48 (42%) high- dose animals. The tumor incidence in male mice from this study is sum- marized in Table 9-19. Of the females, none of the 18 controls, 9 of 48 ~ 19%) low-dose mice, and 17 of 50 (34%) high-dose mice had hepatocellular adenomas; hepa-

372 DRINKING WATER AND HEALTH TABLE 9-19 Tumor Incidence in Male Mice Fed Nitrofen in the Dieta Tumor Dose, ppm Tumor Animal Sex Site (mg/kg/day) Rates B6C3F~ mouse Male Liver 0 1/20 3,000 (132) 3 1/49 6,000 (264) 40/48 aBased on data from NCI, 1979. tocellular carcinomas were observed in none of the 18 controls, in 5 of 48 (10%) low-dose females, and in 13 of 50 (26%) high-dose females. The incidences of hepatocellular adenomas and carcinomas in male and female mice were statistically elevated for the low-dose and high-dose groups (NCI, 19791. Groups of 6-month-old beagle dogs, two of each sex per group, received diets containing 0-, 20-, 200-, or 2,000-ppm concentrations of nitrofen for 2 years. In the 2,000-ppm group, the liver-to-body-weight ratio was significantly increased at termination. Otherwise, there were no observed effects on survival, general health, body weight gain, food consumption, hematologic values, or urine tests conducted at 3-month intervals through- out the study. Furthermore, no histopathological lesions were found in any treatment group. The no-observed-effect level (NOEL) was 200 ppm (5 mg/kg/day) (Ambrose et al., 197 11. On the basis of results from the two NCI studies (1978c, 1979) and the absence of data on humans, IARC (1983, pp. 271-282) concluded that there was sufficient evidence to classify technical-grade nitrofen as a carcinogen in experimental animals. Carcinogenic Risk Estimate In the NCI (1979) study, dose-related increases in the incidence of liver tumors were seen in mice of both sexes, but to a greater extent in males (Table 9-191. These tumor incidence rates were then used to make statistical estimates of both the lifetime risk and an upper 95% confidence limit for lifetime risk for adults weighing 70 kg (see Table 9-201. The risk estimates are expressed as a probability of developing cancer after a lifetime daily consumption of 1 liter of water containing nitrofen at a concentration of 1 ~g/liter and are based on the generalized multistage model for carcinogenesis described in Chapter 8. No risk estimates were calculated for children. If children weighed 10 kg and consumed 1 liter of water per day, their risks would be higher than those of adults because of their higher ratio of water consumption to body weight.

Toxicity of Selected Contaminants 373 TABLE 9-20 Carcinogenic Risk for Nitrofena Estimated with the Generalized Multistage Model Estimated Human Sex Lifetime Riskb Animal B6C3F~ mouse Male 4.4 x 10-5 Upper 95% Confidence Estimate of Lifetime Cancer Riskb 5.6 x 10-5 aBased on data from NCI, 1979. bAssuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ,ug/liter. Developmental Effects lithe teratogenic properties of nitrofen have been examined in an extensive series of studies published over the last 10 years. Those conducted up to 1983 have been reviewed by Burke Hurt et al. (19831. The majority were undertaken with the intent of examining the underlying mechanism of teratogenic activity. Insofar as the exposure levels used in these studies were usually high, they are not well suited for use in risk estimation. These investigative studies are briefly sum- marized here to identify those processes in development most susceptible to disruption by nitrofen. The studies conducted at low, environmentally relevant exposure levels are then examined in detail for application in the risk-assessment process. In a three-generation reproductive toxicity study, Ambrose et al. (1971) first found that nitrofen caused neonatal mortality at dietary levels of 100 ppm and higher. Kimbrough et al. (1974) administered the compound to rats on days 7 through 15 of gestation at 10, 20, or 50 mg/kg/day by gavage and observed that survival of offspring to weaning was reduced at 20 mg/kg and that all animals died at 50 mg/kg/day. Affected pups developed signs of respiratory distress immediately after birth, became cyanotic, and died within an hour. Examination of fetal lungs by electron microscopy revealed an abnormal appearance of Type II cells. Incomplete expansion of alveoli, fibrosis, and immaturity of the alveolar epithelium were observed in treated fetal and neonatal lungs. Stone and Manson (1981) tested whether prenatal exposure to nitrofen accelerated the catabolism of glucocorticoids via induction of cytochrome P450 levels, leading to a depression in glucocorticoid levels and a delay in fetal Jung surfactant synthesis. They found that nitrofen exposure of Sprague-Dawley rats to 20 to 50 mg/kg/day orally on days 8 to 18 of gestation did not induce maternal hepatic mixed-function oxidase activity, unlike findings at identical exposure levels in male and nonpregnant female rats. Cortisone and corticosterone levels in maternal and fetal plasma were unaffected. In addition, fetal lung surfactant synthesis was not altered, as indicated by measurements of total lung phospholipids, i4C-labeled choline

374 DRINKING WATER AND HEALTH uptake in fetal lung lipids, choline phosphotransferase activity, and surface tension values in saline extracts of fetal lungs. They concluded that nitrofen has an inhibitory effect on lung growth but that it does not specifically influence lung surfactant synthesis. Costlow and Manson (1981) found that day 11 of gestation was the most sensitive time for nitrofen-induced neonatal mortality in Long-Evans hooded rats. A range of oral exposures from 75 to 250 mg/kg was given on day 1 1, and the neonatal LDso was found to be 1 16 mg/kg. Studies of maternal toxicity, embryo lethality, and teratogenicity were carried out by exposing the dams to neonatal Lids to LDg9 doses (exposures of 70 to 400 mg/kg) on day 11. No significant treatment effects on maternal weight gain, live fetuses per litter, or major malformations of the fetal skeleton were observed at term throughout the dose range. Fetal body weight was depressed in all treatment groups. Soft tissue examination revealed dose-related incidences of diaphragmatic hernias, hydrone- phrosis, and heart anomalies. Kang et al. (1985) compared the response of different strains of rats Long-Evans hooded, Sprague-Dawley, and virus-antibody negative (VAN) Sprague-Dawley rats exposed orally to nitrofen at 0, 6.25, 12.50, or 25 mg/kg/day on days 6 to 15 of gestation. Abnormalities of the lung (hy- poplasia), kidneys (hydronephrosis), diaphragm (hernia), and heart (aortic arch anomalies, ventricular septal defects, transpositions of great vessels) were observed in fetuses at term. The mean percentage of malformed fetuses per litter was nearly identical across dose levels for each strain of rat. There were, however, substantial differences in the pattern of mal- formations identified within each strain. The Sprague-Dawley and VAN Sprague-Dawley rats responded similarly for all malformations but had significantly higher incidences of diaphragm and lung anomalies than did the Long-Evans hooded rats. Conversely, the Long-Evans hooded rats had significantly elevated levels of kidney anomalies compared with Sprague-Dawley and VAN Sprague-Dawley rats. The frequency of heart malformations was generally low across strains at the dose levels used; significantly elevated incidences occurred at exposure to 25 mg/kg/day and above. These results suggest that although potency (total percentage of fetuses malformed) of nitrofen is similar in different strains of rats, there are marked strain differences in the embryonic target organs affected. A series of studies have indicated that nitrofen alters thyroid function in adult mice and, after prenatal exposure, in offspring. L. E. Gray, Jr., et al. (1982) found that maternal exposure of mice to nitrofen at 6.25 to 200 mg/kg/day on days 7 to 17 of gestation resulted in altered development of the reproductive system and the Harderian glands of offspring effects that were attributed to alterations in the hypothalamic-pituitary-thyroid axis during the perinatal period. Gray and Kavlock (1983) determined that

Toxicity of Selected Contaminants 375 serum thyroxine (T4) levels were significantly reduced after intraperitoneal exposure of nonpregnant mice to nitrofen, whereas triiodothyronine (T3) levels were unaffected. Manson et al. (1984) examined the influence of nitrofen exposure on pituitary-thyroid function in nonpregnant, pregnant, and fetal rats and attempted to relate alterations in thyroid hormone status to induction of birth defects. In adult thyroparathyroidectomized (TPTX) female rats, nitrofen exposure at 15 and 30 mg/kg/day for 2 weeks resulted in significant suppression of thyroid-stimulating hormone (TSH) levels. When a single dose of 250 mg/kg was administered to euthyroid rats, a depression was observed in the release of TSH after a thyrotropin-releasing hormone (TRH) challenge. Pregnant euthyroid rats given a single dose of 250 mg/kg on day 11 had significantly depressed TSH and T4 levels, and fetal T4 levels were markedly depressed at term. Coadministration of T4 and nitrofen to TPTX pregnant rats resulted in a 70% reduction in the frequency of malformed fetuses compared with nitrofen exposure alone. Heart malformations, in particular, were prevented by coadministration of nitrofen with T4. Competitive displacement studies in radioimmunoas- says for T4 and T3 indicated that a nitrofen metabolite (4-aminophenyl 2,5-dichloro-4-hydroxyphenyl ether) competed with I-labeled T3 for antibody binding, whereas the parent compound and six isolated metab- olites failed to compete with i25I-labeled T4 or T3 for antibody binding. These results were interpreted to indicate that the teratogenicity of nitrofen is mediated, at least in part, by alterations in maternal or fetal thyroid hormone status, and that it may be due to a premature and pharmacologic exposure of the embryo to a nitrofen-derived, T3-active metabolite. The studies described thus far involved oral exposure of rodents during pregnancy and examination of the fetuses at term for malformations. The lowest dose found to cause a significant elevation in major malformations in term fetuses was 6.25 mg/kg/day given on days 6 to 15 of pregnancy (Kang et al., 19851. Investigators at the Environmental Protection Agen- cy's Health Effects Research Laboratories have conducted an extensive series of studies on the postnatal sequelae of prenatal exposure to nitrofen. L. E. Gray, Jr., et al. (1982) administered 100 mg/kg/day by gavage to mice on days 7 to 17 of gestation. Dams delivered spontaneously, and offspring were monitored up to 100 days of age. The age of eye opening was significantly delayed in the offspring of treated mice. The height of the palpebral fissures and the weight of the Harderian glands were sig- nificantly depressed at 100 days of age. The absence or reduction of the Harderian glands were only apparent after day 14 of neonatal life, at which time they undergo rapid growth. In the rat, growth of the Harderian glands after eye opening is correlated with an increase in thyroxine secretion. A more extensive examination of the Harderian gland effect was carried out by L. E. Gray, Jr., et al. (19821. These investigators monitored

376 DRINKING WATER AND HEALTH postnatal development in CD1 mice following oral exposure to doses of 0, 6.25, 12.5, 25, 50, 100, 150, and 200 mg/kg/day on days 7 to 17 of gestation. Major malformations and neonatal deaths occurred at 150 and 200 mg/kg/day growth rates were substantially retarded at 12.5, 25, 50, and 100 mg/kg/day. The Harderian glands were reduced or absent in 97% of offspring dosed at 100 mg/kg/day, in 65% at 50 mg/kg/day, and in 4% at 25 mg/kg/day. Gland weights were reduced at all dosages, including the 6.25-mg/kg/day level. Lung and liver weights were also significantly depressed at 110 days of age in offspring exposed prenatally to 6.25 ma/ kg/day. Kavlock and J. A. Gray (1983) examined the effects of prenatal nitrofen exposure on morphometric, biochemical, and physiological aspects of adult renal function. Sprague-Dawley rats were given oral doses of 0, 4.17, 12.5, and 25 mg/kg/day on days 8 to 16 of pregnancy. Nitrofen reduced neonatal survival, primarily because of diaphragmatic hernias, at 12.5 and 25 mg/kg/day. Decreases in body weight, kidney weight, renal protein content, and glomerular number were also observed at these ex- posure levels. Total urine production was suppressed by 12.5 and 25 ma/ kg/day. Pups in the 4.17- and 12.5-mg/kg/day groups failed to concentrate their urine to the same extent as control pups, as evidenced by significantly lower hydroponic urinary osmolalities. The results at 4.17 mg/kg/day were questionable, however, given the range of variability in control animals. There have been additional studies exploring biochemical end points of organ differentiation (Kavlock et al., 1982), morphological and functional defects in the rat kidney and Harderian gland (Kavlock and L. E. Gray, Jr., 1983), and postnatal lung function in the rat (Raub et al., 19831. In these studies, however, the postnatal function deficits were obtained with prenatal exposure levels higher than 6.25 mg/kglday. Developmental Risk Estimate In a study by Weatherholtz et al. (1979), dermal exposure was found to be more effective than oral exposure in reducing neonatal survival. Sprague-Dawley rats were treated percuta- neously with 0-, 0.3-, 3.0-, or 30-mg/kg/day doses of nitrofen (formulated as the emulsifiable concentrate of TOK E-25 containing 25% active in- gredient, xylene, and surfactants) on days 6 through 15 of pregnancy. Similarly treated groups of pregnant females were allowed to deliver their litters naturally, and pups were observed for 4 days postpartum. Major malformations of the diaphragm, accompanied by displacement of lungs and heart and ectopic testes, were observed in the 30-mg/kg/day treatment groups. Neonatal survival to day 4 was significantly reduced in litters treated prenatally with 3 or 30 mg/kg/day. The 0.3-mg/kg/day dose was the NOEL.

Toxicity of Selected Contaminants 377 Costlow et al. (1983) conducted a teratology study to demonstrate dose- response relationships and to establish a NOEL with dermal exposure. Nitrofen was administered dermally to Sprague-Dawley rats on days 6 through 15 of gestation. The compound was prepared as an aqueous dilution of the emulsifiable concentrate and was administered at 0- (water), 0- (solvent), 0.3-, 0.6-, 1.2-, and 12.0-mg/kg/day doses of active ingre- dient. The dams were allowed to deliver their litters naturally, and the offspring were observed for up to 149 days postnatally. Necropsies were performed on dead or moribund animals. Randomly selected offspring were necropsied on postnatal day 42 and the remaining offspring on days 146 to 149. At 12 mg/kg/day, diaphragmatic hernias were found in 95% of the pups found dead at birth, in 17% of the offspring sacrificed on day 42, and in 21% of those sacrificed on days 146 to 149. Survivors in this group also had increased incidences of missing or reduced Harderian glands and slight to severe dilation of the kidneys. In addition, the weights of the thyroid gland, liver, lung, and gonads were decreased in animals surviving prenatal maternal exposure to 12 mg/kg/day. Slight to moderate dilation of the kidneys was observed in all dose groups, including the 0.3-mg/kg/day group, and the severity and incidence increased with dose at day 42 and at days 146 to 149. A NOEL was not demonstrated experimentally but was calculated using the kidney dilation data. The proportion of fetuses within a litter with any degree of renal pelvic dilation was determined for each sex. These data were transformed using an arcsin transformation and plowed against log dose using regression analysis. A weighting factor was used to control for litter size, and the best fit of the model arcsin (Y) = Be + Bi log (dose) was used to predict the NOEL, where Y was the proportion of affected pups in a liner and Be and Bi were constants. For male offspring, the NOEL was estimated to be 0.28 mg/kg/day for exposure on days 6 to 15 of gestation with a 95% confidence interval of 0.12 to 0.34. For female offspring, the NOEL was calculated to be 0.17 mg/kg/day with a confidence interval of 0.05 to 0.27 (Costlow et al., 19831. Unger et al. (1983) administered 0, 0.15, 0.46, 1.39, 4.17, 12.50, and 25 mg/kg/day orally to Sprague-Dawley rats from day 7 of pregnancy through day 15 of lactation to identify dose-response relationships based on pup survival and the incidence of diaphragmatic hernias. Any cyanotic, moribund, or dead pups were placed in fixative and examined for internal malformations by the Wilson technique. On day 29 postpartum, one male and one female per surviving litter were randomly selected, sacrificed, and examined for gross internal abnormalities. At the dose levels used in this study, nitrofen exerted no adverse ma- ternal effects other than nonsignificant reduction in weight gain during

378 DRINKING WATER AND HEALTH TABLE 9-21 Mean Litter Frequencies in Diaphragmatic Hernias in Rats Gavaged with Nitrofena Affected Affected Dose Pups Litters Mean Litter (mg/kg/day) N % N % Frequency 0 0/31 0 0/19 0 0 0.15 2/34 5.88b 1/17 5.88 2.9 0.46 0/21 0 0/14 0 0 1.39 4/32 12.50b 2/16 12.50b 10.9 4.17 5/37 13.51b 4/30 20.00b 17.5 12.50 19/35 54.29b 11/14 78.56b 5l.2b 25.00 32/42 76.19b 6/6 loob 7l .0b aBased on data from Unger et al., 1983. bSignificantly different (p s 0.05) from concurrent control. pregnancy and significantly decreased body weights of lactating dams on days 5 and 8 post partum at the two highest dose levels. The percentage of viable implantation sites, the number of live pups per litter, birth weights, and pup survival were decreased in animals receiving the two highest dose levels. Organ weights of pups selected randomly at weaning were lowered at dose levels ' 4.17 mg/kg/day, the major effect occurring in weights of the Harderian gland. Hydronephrosis was sporadically ob- served in offspring from treated groups and did not appear to be dose related. The incidence of diaphragmatic hernias was significantly different from controls at doses of 1.39 mg/kg/day and higher. The mean litter frequencies of diaphragmatic hernias in pups that died or were sacrificed are given in Table 9-21. This study is limited in its usefulness in identifying a NOEL or a LOEL because only dead or dying pups and two offspring per litter were examined for malformations. This would most likely lead to an underestimation of the true malformation rate insofar as examination was weighted toward identification of lethal malformations. Many of the anomalies associated with prenatal exposure to nitrofen are not lethal. Hydronephrosis, isolated ventricular septal defects, aortic arch anomalies, and even small diaphrag- matic hernias, especially if they are right-sided, have been observed in surviving offspring after prenatal exposure to nitrofen. The decision to perform visceral examinations on dead and dying pups alone probably increased the detection of lethal malformations, e.g., diaphragmatic her- nias, and the decision to examine just two offspring from surviving litters probably led to an underreporting of nonlethal malformations. Conse- quently, the 0.15-mg/kg/day dose should be selected as the LOEL, despite the fact that some outcomes at this level were not found to be significantly

Toxicity of Selected Contaminants 379 elevated when different methods of data presentation and analysis were used. In addition, Unger et al. (1983) did not summarize the mean frequency of malformed fetuses among the litters; rather, the mean percentage of fetuses with individual anomalies (diaphragmatic hernias) was presented. Consequently, the impact of malformations at sites other than the dia- phragm, i.e., the kidneys, lungs, and heart, was not evaluated. It is not possible to reconstruct the mean litter frequency of malformed fetuses from these data because an individual fetus can be represented more than once when data are presented on a malformation basis rather than on a fetus basis. This lends further weight to using 0.15 mg/kg/day as the LOEL insofar as there appeared to be elevations in sites other than the diaphragm at this dose. The Unger et al. (1983) study appears to be the most appropriate for use in cross-species extrapolation, since the authors used oral doses, which are most relevant to drinking water contamination. Two different ap- proaches were considered by the committee in making the cross-species extrapolation. In the first, a LOEL of 0.15 mg/kg/day was identified in the Unger et al. (1983) study. The committee felt that a conservative safety factor of 1,000 was appropriate because a LOEL rather than a NOEL was used and also because nitrofen induces a spectrum of major malformations, some of which are lethal to the newborn while others persist and cause irreversible impairment of the offspring. In addition, these malformations occur at nonmaternally toxic exposure levels (see Chapter 2~. Assuming that a 70-kg human consumes 2 liters of water daily and that 20% of the intake of nitrofen is derived from water, one may calculate the suggested acceptable daily intake (ADI), as 0.15 mg/kg low/day x 70 kg x 0.2 _ 0.0011 mg/liter, or 1.1 ,ug/liter. 2 liters/day x 1,000 In the second, for purposes of comparison, the committee decided to fit the affected litter data shown in Table 9-21 using the probit, logit, and Weibull dose-response models. The computations relating to developmental risk, i.e., one or more pups found defective in a litter, were derived from a computer program prepared by John Kovar and Daniel Krewski of Health and Welfare Canada in 1981. The estimates are all based on the assumption that the response is additive in dose. The 95% lower confidence limit doses were developed by com- puting the confidence limits based on the variance of the reciprocal of the dose. The results of the curve-fitting are given in Table 9-22. In previous volumes of Drinking Water and Health, the risk estimates were averaged to yield one composite number. If the data for all models

380 DRINKING WATER AND HEALTH TABLE 9-22 Developmental Risk Estimates for Nitrofen Dose Dose to Produce a 1 x 10-6 Risk of Defective Offspnnga Upper 95% Confidence (mg/kg/day) Estimate of Lifetime Model MLEb 95~o UCLC Developmental Risky Probit 5.0 x 10-3 2.9 x 10-3 6.4 x 10-6 Logit 5.4 x 10-3 3.0 x 10-3 6.1 x 10-6 Weibull 4.0 x 10-3 2.1 x 10-3 8.8 x 10-6 aBased on laboratory animals that produce litters with several pups as opposed to humans who usually produce a single offspring per birth. bMaximum likelihood estimate. CUpper confidence limit. Assuming daily consumption of 1 liter of water containing the compound in a concentration of 1 ~g/liter. in Table 9-22 are averaged, the upper 95% confidence estimate of lifetime developmental risk is 7.1 x 10-6 and an average dose of 2.7 x 10-3 mg/kg/day which is the upper 95% confidence estimate for a lifetime development risk of 1 x 10-6. Assuming that a 70-kg human consumes 2 liters of water daily and that 20% of the intake of nitrofen is derived from water, one may calculate the suggested ADI from this approach as 2.7 x 10-3 mg/kg low/day x 70 kg x 0.2 _ 0.19 mg/liter, or 19 1lg/liter. 2 liters This value then may be compared with the more conservative suggested ADI of 1.1 ~g/liter obtained from the first approach. These two estimates of "safe" levels (assuming 1 x 10-6 lifetime risk provides a "safe" level) are provided to show the contrast in levels computed using two different reasonable estimation procedures. They are not proposed as a basis for setting appropriate levels for nitrofen in drinking water. Because nitrofen has been shown to be a carcinogen, this effect needs to be taken into account. Some data to support computations which include consid- eration of carcinogenicity are given earlier in Table 9-20. It is difficult to estimate human risks from laboratory animal data since the animals examined produce several offspring per litter in contrast to humans, who usually produce only a single offspring per birth. In addition, when litter data are used for extrapolation, the potential for intralitter interactions cannot be reflected in the results, even though it is not clear what biological significance such interactions would have in risk extrap- olation to humans.

Toxicity of Selected Contaminants 3~31 Reproductive Elects In a three-generation reproduction study, male and female weanling rats were given diets containing 0-, 10-, 100-, or 1,000-ppm dietary concentrations of technical-grade nitrogen. After 11 weeks on this regimen, 20 rats of each sex were mated to produce the successive generation. There were no adverse effects on fertility or ges- tation indices and no apparent effects on viability or lactation indices up to the 100-ppm (10-mg/kg/day) exposures. The viability index for rats on the 1,000-ppm (100-mg/kg/day) diet was significantly depressed, and no pups survived beyond the Fo generation in this group. Likewise, viability was markedly depressed in the 100-ppm group in the progeny of the Fo generation. The NOEL was 10 ppm ~ 1 mg/kg/day) (Ambrose et al., 1 97 1 ). In a two-generation reproduction study, Sherman rats were fed dietary levels of technical-grade nitrofen at 0, 20, 100, and 500 ppm, which furnished doses of 1.1 to 1.8 mg/kg/day, 5.2 to 9.2 mg/kg/day, and 26 to 46 mg/kg/day, respectively. Pair-mating was started when the rats had been on the diets for 68 and 200 days to produce the Fir and Fib litters. Offspring were observed through weaning. The survival of offspring was not affected at the 0- and 20-ppm dietary levels. At 100 ppm the survival of offspring to weaning was reduced, and at the 500-ppm level no offspring survived the neonatal period in two breedings of the first generation (Kim- brough et al., 19741. Technical-grade nitrofen was fed to four groups of 25 male Sprague- Dawley rats for 13 weeks at dietary concentrations of 0, 100, 500, or 2,500 ppm. Untreated female rats were mated with treated males, and fertility, gestation, and lactation indices were monitored. Testes, kidneys, and liver weights in the offspring were increased at 500 and 2,500 ppm (50 and 250 mg/kg/day), but histological changes hypertrophy and cy- toplasmic basophilia of centrilobular hepatocytes were restricted to the liver at levels of 500 and 2,500 ppm. There were no effects on fertility, gestation, litter size, weight, or sex ratio in any group. Offspring health and survival to day 35 were unaffected (O'Hare et al., 19831. Results from these reproductive toxicity studies indicate that exposure of female, but not male, rats to dietary levels of 100 ppm and higher for more than 12 weeks results in decreased neonatal survival. This would be equivalent to approximately 10 mg/kg/day, based on average food consumption values for the rat. CONCLUSIONS AND RECOMMENDATIONS Nitrofen is teratogenic and carcinogenic in laboratory animals. Risk estimates were calculated for both the developmental and carcinogenic effects. Two different approaches were used by the committee in esti- mating developmental risks for humans exposed to nitrofen. In the first,

382 DRINKING WATER AND HEALTH a NOEL approach with a conservative safety factor was used to estimate an acceptable daily intake. In the second, the probit, logit, and Weibull low-dose extrapolation models were used to calculate the developmental risk estimates. All three models gave very similar results, varying from 6.1 x 10-6 to 8.8 x 10-6. The estimated lifetime risk and upper 95% confidence estimate of lifetime risk of cancer in humans are based on the generalized multistage model. The higher risk is due to cancer; however, the short duration of exposure in developmental studies, compared with chronic carcinogenicity studies, indicates that the most sensitive effect may be the developmental one. PENTACH LOROPH ENOL CAS No. 87-86-85 RTECS No. SM6300000 Cl Cl Cl~OH > - Cl Cl Pentachlorophenol (PCP) was first reviewed in Volume 1 of Drinking Water and Health (NRC, 1977, pp. 750-7531. The following material, which became available after the 1977 report was prepared, updates and in some instances reevaluates the information contained in the previous review. Also included are references that were not evaluated in the earlier report. Pentachlorophenol is also known by the trade names Penta, Santophen 20, Dowicide 7 and G. Chlorophen, Penchloro, Sinituho, Weedone, San- tobrite, EP 30, Liropren, Lauxtol, Fungifen, Durotox, Thompson's Wood Fix, Term-l-Trol, Permite, Penta-Kil, Pentasol, Penwar, Perotox, Per- macide, Permagard, Permatox, and Chem-tol. PCP's molecular weight is 266.3. It is a solid at room temperature, it has low volatility, and it is stable in aqueous solution with a solubility in water at 20 ppm at 30°C. In surface waters, PCP generally exists as the anion; some microbial and photodegradation may occur (EPA, 1984b). PCP can be expected to sorb to acidic sediment and leaf litter in surface waters (Hiatt et al., 19601. The odor threshold for PCP in humans is 0.857 mg/ml at 30°C (Hoak, 19571. Pentachlorophenol and its salts (mainly the sodium salt) have been ~n commercial use since the 1930s and are the second most widely used pesticides in the country. Total PCP production in 1980 was 46,826,000

Toxicity of Selected Contaminants 383 lbs (21,240 metric tons) (U.S. International Trade Commission, 1981, p. 2311; imports in 1983 were 274,730 lbs (125 metric tons) (U.S. Inter- national Trade Commission, 1984, p. 981. PCP is also a substantial com- ponent of tetrachlorophenol (Schwetz et al., 19741. Commercial PCP is most commonly produced by the direct chlorination of phenol. It is typically composed of 88.4% pentachlorophenol, 4.4% tetrachlorophenol, less than 0.1% trichlorophenol, and 6.2% higher chlo- rinated phenoxyphenols (Schwetz et al., 19741. The nonphenolic com- ponents fall into two chemical classes: polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans. One commercial product tested con- tained the following chlorinated dioxins: 2,3,7,8-tetrachlorodibenzo-p- dioxin, less than 0.05 ppm; hexachlorodibenzo-p-dioxin, 4 ppm; hep- tachlorodibenzo-D-dioxin. 125 nom: and octachlorodibenzo-p-dioxin, 2,500 ppm (Schwetz et al., 19741. 7 — ``—-—~ The typical dibenzofuran content of commercial pentachlorophenol is as follows: hexachlorodibenzofuran, 30 ppm; heptachlorodibenzofuran, 80 ppm; and octachlorodibenzofuran, 80 ppm (Schwetz et al., 19741. Hexachlorobenzene is also found at levels of 400 ppm in commercial pentachlorophenol (IARC, 1 979; Schwetz et al., 1 9781. More than 80% of the PCP produced in this country is used as a wood preservative. The compound may be used alone or in combination with other agents, such as the chlorophenols, 2,4-dinitrophenol, sodium fluo- ride, the bichromate salts, sodium arsenate, or arsenious oxide (EPA, 1984c). It is also used as a herbicide and defoliant. Approximately 12% of the PCP produced is used in making sodium pentachlorophenol- a wood preservative, fungicide, herbicide, and slimicide (IARC, 19791. PCP has been detected in a variety of wildlife and in fish at levels ranging from 0.35 to 26 mg/kg (IARC, 1979; Renberg, 19741. In the State of Michigan, herds of dairy cattle became contaminated with PCP that was used to treat the walls and feed bins of their barns. The PCP levels in 18 of these cows ranged from 58 to 1,136 ~g/kg (Anonymous, 1977; IARC, 19791. Widespread human exposure to PCP occurs as a result of its extensive production, persistence in the environment, and varied applications. Most exposure arises from its use as a wood preservative; however, there is also extensive exposure from use in homes and gardens. There are many reports of occurrence in the food chain, general environment, and work- place as well as in the body fluids of the general population and exposed workers. And there are a number of reports of toxicity in people exposed chiefly in the home. The National Institute for Occupational Safety and Health noted that most workers exposed to PCP are employed in the gas and electric service industries and that hospital workers may also be ex- posed (NIOSH, 1977b).

384 DRINKING WATER AND HEALTH The following paragraphs provide some examples of environmental concentrations. Air PCP concentrations from 0.25 to 7.8 ng/m3 have been detected in the air of two towns (Cautreels et al., 1977; IARC, 19791. In a contaminated home, the air contained PCP concentrations ranging from 0.14 ~g/m3 in a bedroom to 1.2 ~g/m3intheattic(Sangsteret al., 19821. Inawoodpre- servative factory and wood treatment plant, air levels were 0.013 mg/m3 (average) and 5.1 to 15,275 ng/m3, respectively (IARC, 19791. Water PCP has been measured in effluent streams (0.1 to 10 mg/liter) (IARC, 19791; in river water (1.23 ~g/liter) (Wegman and Hofstee, 1979~; and in surface ponds and drainage water near a wood treatment facility (1 to 800 ,ug/liter) (Won" and Crosby, 19811. Concentrations of 98 ppt have been detected in Dade County, Florida, municipal drinking water (Morgade et al., 19801. Levels up to 24 ppm have been detected in wells near a sawmill in Hayfork, California (Litwin et al., 19831. Sediment In Europe, measured concentrations in sediment have been 20-fold higher than those found in surface water (EPA, 1984b; Wegman and van den Broek, 19831. Soil Levels from 0.46 to 69.1 mg/kg have been observed in greenhouse soil treated with PCP (IARC, 1979~. Food In the United States, the estimated average daily intake of PCP from food in 1974 was 0.76 fig (IARC, 19791. Animals Levels ranging from 0.35 to 23 mg/kg in fish and from 58 to 1,136 ~g/kg in cattle have been found in animals contaminated with PCP used to treat wood (IARC, 19791. Humans In nonoccupationally exposed dialysis patients, PCP levels in blood plasma were 15.7 to 15.8 ~g/liter, compared with 15.0 ~g/liter in controls (IARC, 19791. In the general, nonoccupationally exposed pop- ulation, PCP has also been detected in seminal fluid (20 to 70 ~g/kg), fingernails (IARC, 1979), blood, and urine (Dougherty, 19781. In a recent survey, PCP was found in 85% of the urine samples collected. The mean level was 6.3 ~g/liter (Kutz et al., 19781. Adipose tissues from necropsy samples taken from residents of Dade County, Florida, contained PCP concentrations ranging from 10 to 80 ppb (Morgade et al., 19801. PCP was detected in workplace air at 25 factories in which the compound was used to treat wood. The average level was 0.013 mg/m3 (maximum

Toxicity of Selected Contam inants 385 range, 0.004 to 1.0 mg/m31. The level in urine of the workers exposed to these airborne concentrations ranged from 0.12 to 9.68 mg/liter (Arsenault, 1976; IARC, 19791. Worker exposure to PCP at one wood treatment plant over a 5-month period resulted in serum and urine levels of 348.4 to 3,963 ~g/liter and 41.3 to 760 ~g/liter, respectively (IARC, 1979; Wyllie et al., 19751. Plasma levels of 0.02 to 2.4 ~g/ml have been reported in PCP producers and applicators (Zober et al., 19811. Persons exposed to PCP in contam- inated homes had plasma levels of 25 to 660 Igniter. An average of 128.6 + 134 ~g/liter (mean + SD) was seen in controls (Sangster et al., 19821. PCP is registered in the United States as a wood preservative and for agricultural use in seed treatment. In 1978, however, the Environmental Protection Agency (EPA) issued a rebuttable presumption against regis- tration for pesticide products containing PCP and in July 1984 limited their sale and use, also setting restrictions on dioxin levels in PCP- containing products (EPA, 19781. The American Conference of Govern- mental Industrial Hygienists (ACGIH) has established a threshold limit value of 0.5 mg/m3 and a short-term exposure level (STEL) of 1.5 ma/ m3 for dermal exposure (EPA, 1984c). The Occupational Safety and Health Administration's 8-hour time-weighted average is 0.5 mg/m3 (IARC, 19791. In 1980, EPA recommended alternative ambient water quality criteria of 1.01 mg/liter based on PCP's toxicity and 0.030 mg/liter based on the organoleptic properties of PCP. EPA has recently derived 1-day, 10-day, and lifetime health advisories for PCP in drinking water, all of which exceed 1 mg/liter for adults and are less than 1 mg/liter for children. The agency recognized that the odor threshold for PCP is considerably lower (0.857 mg/rnl at 30°C) (EPA, 1984b). METABOLISM Rapid absorption of PCP has been reported in rodents, monkeys, and humans following oral, dermal, or inhalation exposure. In healthy vol- unteers ingesting a single dose of 0.1 mg/kg bw, the average half-life for absorption was 1.3 + 0.4 hours. Peak plasma concentrations occurred after 4 hours (Braun et al., 19781. The major tissue deposits vary somewhat between species. In humans whose deaths were not related to PCP exposure, the liver (containing PCP residues of 0.067 ~g/g), kidney, brain, spleen, and fat (0.013 ,ug/g) appeared to be major deposition sites (Grimm et al., 1981~. In the mouse, the gall bladder is a principal storage site. In the rat, it is the kidney. The metabolism of PCP is generally similar in mammalian species. In rodents, more than 40% is excreted in urine unchanged. The remainder is excreted as tetrachlorohydroquinone and glucuronide conjugates of PCP.

386 DRINKING WATER AND HEALTH In limited studies of humans, PCP, tetrachlorohydroquinone, and PCP- glucuronide have been found in urine (Ahlborg et al., 1974; Braun et al., 1977, 19781. In vivo retention of PCP by lipid-containing tissues may be attributable to conjugation with fatty acids (Leighty and Fentiman, 19821. In all species tested, including humans, PCP is excreted principally in the urine; smaller amounts are found in feces and expired air. Excretion in rodents and humans is apparently biphasic (Bevenue et al., 1967~. The excretion kinetics of PCP vary appreciably, depending on whether the exposure is acute or chronic. In humans, urinary excretion half-lives fol- lowing chronic exposure are significantly longer than after single high- dose exposure (20 days versus 10 hours) (Casarett et al., 19691. Blood and urine PCP concentrations were monitored in 18 workers (Begley et al., 1977) 1 day before taking vacation, on 4 different days during their vacations, and about 50 days after they had returned to work. During the vacation period, PCP levels in blood and urine fell to < 50% of the prevacation values. From these data, the estimated elimination half- life of PCP was calculated to be 19 to 20 days (EPA, 1984b). In vitro studies have suggested the following possible mechanisms of toxicity (which are not necessarily mutually exclusive): Uncoupling of oxidative phosphorylation: Weinbach (1954), A~Thenius et al. (1977), and Gotz et al. (1980) showed that PCP uncouples mito- chondrial phosphorylation in rat hepatocytes, indicating that hepatotoxic effects may be due to an interference with the energy metabolism of the cell. Uncoupling may result from the binding of PCP with mitochondrial protein (Weinbach and Garbus, 1965) or with fatty acids in the mito- chondrial membrane (Leighty and Fentiman, 19821. Disturbance of microsow~al detoxification: PCP induces microsomal en- zymes (Vizethum and Goerz, 19791. However, in vitro studies of rat liver microsomes have shown that PCP inhibits microsomal detoxification enzymes by disturbing electron transport from flavins to cytochromes. This suggests that PCP may synergistically increase the toxicity of many substances, such as polynuclear aromatic compounds (Arrhenius et al., 19771. Perturbation of lipid membranes by PCP: Packham et al. (1981) have shown a correlation with phospholipid perturbation and toxicity (LDso) for a series of compounds. PCP's interaction with a number of other chemicals: The toxicity in Pseudomonas fluorescens was greater when PCP and 2,3,4,5-tetrachlo- rophenol were given sequentially than when PCP alone was given (Trevors et al., 19811. The antioxidant butylated hydroxyanisole (BHA) enhances the toxicity of PCP to P. fluorescens (Trevors et al., 19811. Hexachlo- robenzene (HCB) at 1,000 ppm and 99% pure PCP at 500 ppm admin- istered to female Wistar rats for up to 8 weeks resulted in an increased

Toxicity of Selected Contaminants 387 accumulation of PCP in the liver. PCP also accelerated the onset of hepatic porphyria by HCB (Debets et al., 19801. Pretreatment with PCP inhibits the carcinogenic effect of hydroxamine acids and the hepatoxicity of N- hydroxy-2-acetylaminofluorene (Meerman and Mulder, 1981 ~ . HEALTH ASPECTS Observations in Humans There are numerous reports of toxic effects and deaths due to occu- pational and accidental exposures to PCP and sodium PCP (EPA, 1984b; IARC, 1979; Mercier, 19761. For example, factory workers exposed to PCP at a Winnipeg plant experienced sweating, weight loss, and gas- trointestinal disorders (Bergner et al., 1965) as did herbicide sprayers,in Australia (one person exposed for an extended period suffered hepatic enlargement) (Gordon, 19561. Plant workers in the Federal Republic of Germany reported skin eruptions and bronchitis following "acute" ex- posure (Baader and Bauer, 19511. Other responses to extended occupational exposure to PCP that may be related to contaminants include persistent chloracne and disorders of the liver and nervous system (IARC, 19791. In addition, German workers (applicators and producers) had a significant rise in immunoglobulins, compared with controls, following chronic exposure (Zober et al., 19811. Hematological effects, including atteinte medullaire (bone marrow dam- age), have been observed in workers exposed to products containing 5% PCP and 1% lindane in a petroleum solvent (Catilina, 1981; Catilina et al., 19814. Average serum and urine levels of PCP were higher (about 30% and 50%, respectively) in six exposed workers than in four controls, but no significant differences were seen in the incidence of chromosome aberrations (Wyllie et al., 19751. Repeated dermal exposure to sodium pentachlorophenate may result in dermatitis, systemic intoxication, and, in a limited number of people, an allergic response (EPA, 1984b.) Cases of acute nonfatal poisoning have followed the home or hospital use of products containing PCP. These include a 3-year-old child who, after exposure to PCP in a contaminated water supply, experienced delirium, fever, and convulsions (Chapman and Robson, 19651. Twenty neonates who were exposed to PCP used in laundering diapers and bed linen subsequently de- veloped tachycardia, respiratory distress, and liver changes (Robson et al., 19691. In another study, people were found to have dermatitis after exposure to PCP-treated lumber in their houses (Sangster et al., 19821. Fatal poisonings include a 58-year-old male employed for 1 week as a wood dipper. His death was most probably due to cutaneous absorption

388 DR! NKING WATER AND H EALTH and inhalation of PCP. Autopsy revealed degeneration of the liver and kidney (Bergner et al., 19651. Five fatal cases were reported among her- bicide sprayers in Australia (Gordon, 19561; nine deaths occurred in saw- mill workers in Asia (Menon, 19581. Two of the 20 neonates mentioned above died, and an autopsy indicated that they had degeneration of the kidney and effects on the liver (Robson et al., 19691. Roberts (1963) has reported a fatal case of aplastic anemia in a 21- year-old male who handled treated lumber. More recently, he cited three additional cases of fatal aplastic anemia following exposure to PCP (Rob- erts, 1981~. One of these cases also had Hodgkin's disease. Roberts (1981) also reported several cases of soft-tissue sarcoma, Hodgkin's disease, non- Hodgkin's lymphoma, and other malignant diseases after exposure to PCP, related compounds, and precursors. A series of studies of chronically exposed workers has been conducted in Hawaii. The first involved workers in wood treatment plants and farmers or pest-control operators. Elevation of serum enzyme levels, i.e., serum glutamic-oxaloacetic transaminase (SGOT), serum glutamic pyruvic trans- aminase (SGPT), and lactic dehydrogenase (LDH), and low-grade infec- tions or inflammations of the skin, eye, and upper respiratory tract were found in the exposed groups (Klemmer et al., 19801. In a second study, plasma protein levels were found to be elevated in exposed, as compared with unexposed, workers (Takahashi et al., 1976~. Begley et al. (1977) demonstrated an effect of PCP on the renal functions of 18 workers and an improvement following vacation. No long-term epidemiological studies of carcinogenicity were found by the committee. Observations in Other Species Acute Effects At least 40 LD50 and LD~o studies have been performed, many of them with PCP of undefined purity. Therefore, it is difficult to determine the extent to which effects are attributable to PCP or to con- taminants. Acute exposure of laboratory animals to PCP results in vomiting, hy- perpyrexia, elevated blood pressure and respiration rate, and tachycardia. Oral LDsos ranging from 27 to 150 mg/kg bw have been reported, but no apparent differences in susceptibility have been observed between species at the lower level (Deichmann et al., 1942; Dow Chemical Company, 1969; Schwetz et al., 19781. No interspecies differences were reported for the higher level. SubchroniclSubacute Effects A significant number of subchronic/sub- acute studies have been conducted in rats. Both technical- and analytical- grade PCP were used, allowing attribution of some but not all observed

Toxicity of Selected Contaminants 3139 toxic effects to contaminants in technical PCP. In a study to determine the subchronic toxicity of the compound, PCP was fed in the diet to groups of Wistar rats at concentrations of 0, 25, 50, and 200 ppm for a 90-day period (Knudsen et al., 19741. Female rats receiving PCP at 200 ppm (10 mg/kg/day) had reduced growth rate. Liver weight in the female rats fed 50 and 200 ppm (2.5 and 10 mg/kg) daily was significantly higher. After 6 weeks, male rats fed 50- and 200-ppm concentrations of PCP had elevated hemoglobin and hematocrit values, whereas at 11 weeks he- moglobin and erythrocytes were significantly reduced in the same groups of animals. No PCP-related effects were observed in animals fed 25 ppm (1 .25 mg/kg/day) (Knudsen et al., 1974; NRC, 19771. In another experiment, male rats received technical or pure PCP in doses of 50 mg/kg/day for a 90-day period (Kimbrough and Linder, 19784. Both PCP formulations caused an increase in liver weight. More severe histopathological changes occurred in the livers of rats given the technical PCP than in those given the pure PCP (NRC, 19771. In a 90-day study, increased liver and kidney weights, elevated serum alkaline phosphatase, and depressed serum albumin levels were observed in Sprague-Dawley rats that consumed technical-grade PCP at 3, 10, and 30 mg/kg/day (Johnson et al., 1973~. When a sample of PCP containing substantially reduced amounts of dioxins was fed to rats, no adverse effects were found at 3 mg/kg/day. In animals receiving pure PCP, kidney and liver weights were elevated at 10 and at 30 mg/kg/day, but no adverse toxicological effect was found in animals receiving 3 mg/kglday (NRC, 19771. Greichus et al. (1979) observed an increase in liver weights of pigs fed purified PCP in doses of 10 and 14 mg/kg bw daily for 30 days. Subcutaneous administration of sodium pentachlorophenate (purity un- known) to three groups of six rabbits each for 60 days (at doses of 13.8, 27.5, or 68.8 mg/kg bw) caused a variety of effects, including secondary anemia, leukopenia, and numerous lesions in the brain and spinal cord. Doses were one-twentieth, one-tenth, or one-fourth the minimum lethal dose (275 mg/kg) (McGavack et al., 19411. In vivo subchronic exposures of Wistar rats exposed to technical-grade PCP at 30 mg/kg bw resulted in alterations of mitochondria and nuclei of hepatocytes (Fleischer et al., 19801. Chronic E~ects In a chronic study, alterations in liver morphology were observed at all doses in rats ingesting from 20- to 500-ppm concen- trations of technical PCP over an 8-month period (Kimbrough and Linder, 1978~. By companson with pure PCP, more severe histopathological changes occurred in livers of rats given technical-grade PCP. With pure PCP, effects were not observed at the lowest dose.

390 DRINKING WATER AND HEALTH Goldstein et al. (1977) administered both pure and technical-grade PCP to the diet of female Sherman rats for more than 8 months. Hepatic porphyria occurred at 100 and 500 ppm, whereas increased hepatic aryl hydrocarbon hydroxylase (AHH) and glucuronyl transferase (GT) activ- ities were observed at 20-ppm concentrations of technical-grade PCP. By contrast, pure PCP at 500 ppm had an effect only on GT activity and body weight. Thus, the porphyria and other major liver changes induced by technical PCP are apparently due to contaminants, probably the chlo- rinated dibenzo-p-dioxins rather than the PCP. In addition, Schwetz et al. (1978) observed decreased body weight (females only), increased urine-specific gravity (females only), and in- creased SGPT activity (both sexes) in 27 male and 27 female Sprague- Dawley rats fed 30-mg/kg bw doses of purified PCP for 2 years. Pig- mentation of liver and kidneys was observed in females receiving 10 or 30 mg/kg bw daily and in males receiving 30 mg/kg bw each day. No- effect levels for females were 3 mg/kg/day; for males, 10 mg/kg/day. In a 160-day study, cattle fed 20-mg/kg doses of technical PCP for 42 days, followed by 15 mg/kg/day for the remainder of the study, had decreased weight gain, progressive anemia, and immune effects. Only minimal adverse effects were observed after exposure to analytical-grade PCP (McConnell et al., 19801. Immunotoxicity Kerkvliet et al. (1982a,b) observed adverse effects on the immune system of B6 male mice given technical-grade and pure PCP at doses of 50 or 500 ppm of the diet for 10 to 12 weeks. The technical- grade PCP at 50 and 500 ppm increased both the incidence of transplanted tumors and susceptibility to virus-induced tumors. It also reduced T-cell cytolytic activity and caused a dose-dependent delay and suppression of peak splenic antibody production and serum antibody titers. The only effect of pure PCP at 500 ppm was to enhance development of splenic tumors (which were actually virus-induced tumor cells) following chal- lenges with MSV/MB virus. By contrast to the enhancement of tumor susceptibility in PCP-exposed mice, PCP exposure did not significantly alter the susceptibility of animals to infectious virus-induced mortality. Kerkvliet et al. (1982a) also observed immunotoxicity of technical- grade PCP in female Swiss-Webster mice. Animals were given diets con- taining technical-grade PCP for 8 weeks at 0, 50, 250, and S00 ppm (0, 7.5, 37.5, and 75 mg/kg). Immunosuppression was seen at the lowest dose. These effects were not found after administration of pure PCP. Significant effects have also been found in the immune system of Sprague-Dawley rats (Exon and Koller, 1983b) and cattle (McConnell et al., 19801. Exon and Koller (1983b) fed pentachlorophenol (97%) in feed to female Sprague-Dawley rats from weaning until 3 weeks post partu-

Toxicity of Selected Contaminants 391 rition. Offspring from each treatment regimen were continued on PCP treatments until 13 weeks of age. Delayed hypersensitivity, suppression of humoral immunity, and increased phagocytosis by macrophages were observed at all dose levels (5, 50, and 500 ppm, i.e., 0.5, 5, and 50 ma/ kg) in both pre- and postnatally exposed animals. Neurotoxicity McGavack et al. (1941) observed effects on the central nervous system in rabbits after 60 days of exposure to subcutaneous doses of 5%, 10%, and 25% of the minimum lethal dose (275 mg/kg bw). Nervous system lesions were seen in all dose groups. Neurochemical effects were observed in 30 male Wistar rats given 20-mg/liter concen- trations of technical-grade PCP in drinking water for 3 to 14 weeks. Thirty controls were also studied. Animals were sacrificed at 3 to 18 weeks after the experiment began. The main effects seen in the rat brain were transient biochemical effects (e.g., altered acid proteinase activity, superoxide dis- mutase, and glial glutathione content) (Savolainen and Pekari, 19791. Mutagenicity There are only limited data from in vivo genetic testing of PCP. In Drosophila melanogaster, there was no increase in sex-linked recessive lethal mutations (Fahrig et al., 1978; Vogel and Chandler, 19741. In the mouse spot test, there was evidence of definite but weak mutagenic activity (Fahrig et al., 1978~. In vitro testing produced negative results in numerous bacterial mutation assays with the exception of a positive Bacillus subtilis rec assay (Waters et al., 19821. In addition, positive results were observed in a yeast assay for forward mutation (Saccharomyces cerevisiae) (Fahrig et al., 19781. Chromosome abnormalities were seen in Vicia fabia seedlings (Amer and Ali, 19691. No increases in chromosome aberrations were observed in workers with elevated serum and urine levels of PCP (Wyllie et al., 19751. Interpretation of many of these studies is somewhat handicapped by in- adequate reporting of dosages, ranges of background mutation, and other factors (Williams, 19824. Carcinogenicity PCP (technical-grade and commercial-grade) is cur- rently under test by the National Toxicology Program (NTP, 1985b). No statistically significant differences between experimental animals and controls were observed in bioassays for cancer in which commercial-grade PCP was given to mice (18 per group, about 20 months exposure by gavage) (BRL, 1968a, p. 3931; in Sprague-Dawley rats given dietar,~ doses of commercial- grade PCP (27 per group, 22 to 24 months) (Schwetz et al., 19781; or in Wistar rats given both purified and technical-grade PCP by subcutaneous injection for 40 weeks (Catilina et al., 19811. In another study, however, the incidence of hepatomas in one of two strains of mice t(C57BL/6 x C3H/

392 DRINKING WATER AND HEALTH Anf)F~] tested was significantly increased following subcutaneous injection of commercial-grade PCP (IARC, 1979; BRL, 1968a). Commercial-grade PCP was not found to be a promoter of skin carcinogenesis in mice initiated with dimethylbenzanthracene (DMBA) (Boutwell and Bosch, 19591. As mentioned earlier, PCP enhances the transplacental carcinogenicity of ethyl- nitrosourea (Exon and Koller, 1982, 1983a). The carcinogenicit,v studies conducted to date do not provide a basis for complete evaluation because of limitations, including inadequate duration, lack of concurrent testing in two species, and small numbers of animals (Williams, 19821. Developmental Elects Purified and commercial PCP administered orally to rats at daily doses of 5 to 50 mg/kg bw at various intervals during days 6 through 15 of pregnancy resulted in dose-related toxicity in the embryo and developing fetus. Early organogenesis (days 8 to 11) was the most vulnerable stage. Unlike other effects of PCP, these effects were somewhat greater with purified PCP than with the commercial product. At 5 mglkg/ day administered on days 6 through 15 of gestation, commercial PCP had no adverse effects, but pure PCP caused a statistically significant increase in the incidence of delayed ossification of the skull. No other effects on emb~yonal and fetal development were noted (Schwetz et al., 19741. A single 60-mg/kg bw oral dose of purified PCP was given to pregnant Charles River CD strain rats on days 8, 9, 10, 11, 12, and 13 of gestation. Treatment on days 9 and 10 had the greatest effect on fetotoxicity (Larsen et al., 19754. Mice given 25-mg/kg bw doses of commercial PCP in dimethyl sulfoxide by subcutaneous injection on days 6 through 14 of gestation (BRL, 1968b, p. 150) developed no teratogenic effects. Pregnant Syrian golden hamsters given daily oral doses of PCP (unspecified purity) ranging from 1.25 to 20 mg/kg from days 5 to 10 of gestation experienced an increase in fetal deaths and resorptions. The no-effect level was 2.5 mg/kg/day (Hinkle, 19731. Daily dietary exposure of Sprague-Dawley rats to 0, 3, or 30 mg/kg bw for a 62-day period before mating and continuing through lactation resulted in significant decreases in neonatal body weight, percentage of live births, and survival of pups at a dose of 30 mg/kg/day (Schwetz et al., 19781. Exon and Koller (1982, 1983a,b) administered PCP (0, 5, 50, or 500 ppm) to Sprague-Dawley rats in the diet beginning with the rats' own weaning through the weaning of their pups. They observed significant effects on the immune system (as indicated by decreased antibody titers, decreased delayed hyper- sensitivity to oxazolone, and increased peritoneal macrophage numbers) and reduced ethylnitrosourea-induced transplacental carcinogenesis.

Toxicity of Selected Contaminants 393 CONCLUSIONS AND RECOMMENDATIONS Inadequate characterization of PCP preparations used in toxicity ex- periments has led to uncertainties about the extent to which observed effects are attributable to PCP or to contaminants, including dioxins. The metabolism of PCP is generally similar among mammalian species. PCP can be asborbed through skin, by inhalation, and by ingestion. Chronic exposures result in longer retention of PCP than do acute exposures. PCP interacts with other chemical substances that may increase or decrease its toxicity. In humans, case reports indicate neurotoxicity, immune system effects, liver and kidney damage, and hematological disorders. There are also reports relating aplastic anemia and malignancy (Roberts, 1963) to PCP exposure. Epidemiological studies indicate alterations in enzyme systems and renal function and decreased resistance to disease associated with chronic exposure. No reports of epidemiological studies of possible car- cinogenicity were found by the committee. Studies in experimental animals have demonstrated effects from acute, subchronic, and chronic exposure. These include damage to liver, kidney, and central nervous system; hematological and immune system effects; and fetotoxicity. Pure PCP was more fetotoxic than technical-grade PCP. The results of mutagenicity testing are mixed. PCP gave weak results in the mouse spot test, was positive in yeast and Bacillus subtilis, but was negative in several other short-term in vitro assays. Two-year chronic bioassays in two species have not yet been completed. Therefore, carcinogenicity bioassays to date do not provide an adequate basis for evaluation. Thus, EPA based its recent risk assessment on the National Cancer Institute bioassay of certain PCP contaminants, the hexachlorodi- benzo-p-dioxins (EPA, 1984b; NCI, 19801. A PCP chronic bioassay is cur- rently being conducted by the National Toxicology Program COP, 1985b). PCP should be reevaluated when the results of that bioassay become available. Developmental toxicity is an end point of considerable current concern. Animal and human studies suggest that the fetus and neonate are partic- ularly susceptible. In the one-generation reproduction study by Schwetz et al. (1978), 3 mg/kg/day was administered in the diet 62 days prior to mating, during 15 days of mating, and subsequently throughout gestation and lactation. This may be regarded as a no-observed-effect level (NOEL), although there was a trend toward decreased neonatal weight at this dose. This is consistent with other studies shown in Table 9-23 showing NOELs ranging from 1 to 3 mg/kg/day. A previous Safe Drinking Water Committee calculated an acceptable daily intake (ADI) of 0.003 mg/kg by using an uncertainty factor of 1,000

394 o i_ 4 - Cat o _) sit o _' U) o o By o Cal V) ° o 1 t-4 ~ o C) Pa ~:5 Ct Lo ~ _` ~ o ° sit z o ~ ~ A Cat . 'I flu o Cal US ·— Ed ~ _ Ce _ · o ~ ~ ce c!S _ 3> _ ~ ~ 4) t_ o [_ ~ oo C<S ~ ~ ~ —~ ~ — 3 ~ ~ c~ c~ ~ ~ ~ o ~ =: o~ ~ ~ ~ ~ ~ ~ ~ ~ .s ~ ~° ~ ~ =~ ,3., ~ ,a~ ~4 C~ ~— e ~ e , y -e · .~ e e e O x 2 ' A ~ ~ — ~ O C ~ z 7- . ~! ~' .e. ~ C~ ~ . . ~ ~ ~ ~ O ~ O ~ =°° ~ == == E ~ ~ E ~ E o ~ 8 o ~ o ;^ 8 ~ ~ ~ 8 ° 8 I ~ I ~ ~ ~ ~ ~ ~ 3 ~ ~ ~ o 0 0 0 0 ~ o 0 0 tn ~ c 3 ~ :^ ~ ~ 3 =~ =o ~ ~ ~ a~ ~ ~ ~ oO Oo -_ ~ 4_ ~ ~ ~ ~ ~ ct ct ct ct a~ C>, ~ C) ._ ._ ._ =; ~ ~

395 ^ ~ ~ ~ oo ~ of ^ ^ Ct · · ~ ~ ~ ~ ~ (¢ Ct _ ~ ~L, ~ 0, ~ 0~ ° ~ 2 ~ ~ ~ ~ ~ ~ oo ~ cot 3 ~ 3 ~ ~ t~ ·— ~ — ~ ~ ~ ~ 3 C) ~ ~ ~ i= ~ . ~ o o ~ ~ V) ~ ~ ~ ~ V ~ U. . ~ ~ Cal ~ , ' ~ ' _ ~ ~ ' c ' o~ 3 2 —~ ~ r ~ E Z Z ~ ~ ~ ~ Z Us ~ . . Us Vie ~ ~ _ Cal Cal o.e ~,4 on ~ o ~ ~ ~ PI O s=^ s ~> I =,,0 0 0 0 e ~^ O O O _ O O O O O by O O O ~ ~ ~ ~ ~ O ~ 0 ~ 04 ~ ~ ~ ~a ~ ~ ~ ~ ce ~ ~ ~ ~ oo oo oo oo ~ ~ V~ C) _ _ ~ ~ _ ~ _ · _ C\: Ce ~ Ct C<S Ct Ct Ce ~ ~ ~ := ~ ~ ~ ~ ~ C~ oc C) ._ _ _ .= .= ;^ cn '_ 0,0 ~ ~ V~ . ~ ~ ' ~ C~ ~ . ~ oc e~, cS ~ ~ ~

396 DRINKING WATER AND HEALTH (NRC, 19771. Using this as a basis, and assuming that a 70-kg human consumes 2 liters of water daily, which contributes 20% of total intake, the present committee estimated a chronic suggested no-adverse-effect level (SNARL) as: 3 mg/kg/day x 70 kg x 0.2 0.021 mg/liter, 1,000 x 2 liters or 21 ~g/liter. This value is consistent with current NOELs for pure PCP. A SNARL may also be estimated for a 10-kg child consuming 1 liter of water daily, which contributes 20% of total intake: 3 mg/kg/day x 10 kg x 0.2 _ 1,000 x 1 liter _ 0.006 mg/liter, or 6 1lg/liter. The committee noted that the toxicity of PCP is increased by impurities contained in the technical product. For example, the NOEL for pure PCP is 3 mg/kg/day (Johnson et al., 19731; however, the NOEL for technical PCP is 1 mg/kg/day (Goldstein et al., 1977; Kimbrough and Linder, 1978), indicating increased toxicity due to impurities. A SNARL for an adult may also be calculated for technical PCP as: 1 mg/kg/day x 70 kg x 0.2 0.007 mg/liter, 1,000 x 2 liter or 7 1lg/liter. Likewise, a NOEL of 1.25 mg/kg/day for commercial PCP (Knudsen et al., 19-14) could be used to calculate a SNARL of 9 ,ug/liter, which would fall in the range between the SNARLs for technical and pure PCP. The committee recommends the lower SNARLs for exposure to com- mercial- and technical-grade PCP. When the carcinogenicity data from the NIP bioassay become available, PCP should be reevaluated. TRICHLORFON Dimethy' (2, 2, 2-trichIoro- ~ -hydroxyethyI)phosphonate CAS No. 52-68-6 RTECS No. TA0700000 O H 11 1 CH3—O—P—C—CC13 1 1 O OH CH3 Trichlorfon is used both as a pesticide and as a chemotherapeutic agent for schistosomiasis. Trade names for the insecticides include Dylox,

Toxicity of Selected Contaminants 397 Dipterex, and Clorfos. The pharmaceutical compound is usually called metrifonate, and commercial names include Bilarcil and Metriphonate. In aqueous solutions, acid conditions promote the degradation of tri- chlorfon to dimethylphosphonic acid and trichloroethanol (IARC, 1983, pp. 207-231), whereas neutral-to-alkaline aqueous solutions result in a rearrangement of the compound to dimethyl(2,2-dichloroethenyl)- phosphate (Dichlorvos) (Dedek et al., 1969; Metcalf et al., 19591. Ultra- violet light is also reported to degrade trichlorfon. The World Health Organization (WHO) reported that the half-life of trichlorfon sprayed on green plants is 1 to 2 days (WHO, 1972~. The International Agency for Research on Cancer (IARC) reported that trichlorfon persists in soil for up to 2 weeks (IARC, 1983, pp. 207-231), while WHO cited a technical report stating that complete loss (below limits of detection) of a 10-ppm concentration of trichlorfon was achieved in 15 to 112 days, depending on soil type. Trichlorfon showed a "low" tendency to move to ground- water from soil (WHO, 19721. Most data on human exposure to trichlorfon come from reports on use of the compound as an antischistosomal drug (Lebrun and Cerf, 19601. For this purpose, the compound is normally ingested in doses ranging from 5 to 12.5 mg/kg (Holmstedt et al., 19781. Originally, trichlorfon was given on consecutive days, but current regimens require three doses to be given over 2 to 4 weeks (Jewsbury et al., 1977; Reddy et al., 1975~. Humans have also been reported to have been exposed occupationally (van Bao et al., 1974) or in suicide attempts (Hierons and Johnson, 1978; Senanayake and Johnson, 19821. METABOLISM After humans were exposed to a 7.5- to 10-mg/kg dose of trichlorfon, presumably by the oral route, peak plasma levels of the compound were detected in plasma in 1 hour (Nordgren et al., 19814. Comparable LD50s for experimental animals given trichlorfon orally and parenterally (500 and 250 mg/kg, respectively) led Holmstedt et al. (1978) to speculate that the rate of absorption is similar via these routes. By this reasoning, the high LDso of dermally administered trichlorfon (greater than 2,800 ma/ kg in rats, according to Edson and Noakes, 1960) would suggest that the compound is slowly or inefficiently absorbed from the skin. In rats intravenously injected with ~4CH3-labeled trichlorfon, Dedek and Lohs (1970) found radioactivity in liver, lung, kidney, heart, spleen, and blood. The inhibition of esterases in the spinal cord and brain of hens suggested that trichlorfon or an active metabolite also enters the central nervous system (Hierons and Johnson, 19781.

398 DRINKING WATER AND HEALTH Arthur and Casida (1957) suggested that the primary fate of trichlorfon is hydrolysis of the phosphonate bond to form 2,2,2-trichloroethanol, which undergoes glucuronidation, and dimethyl phosphate. Another major reaction is demethylation of the parent compound, which is probably mediated by glutathione. As reviewed by Holmstedt et al. (1978), non- enzymatic rearrangement of trichlorfon to dichlorvos is only a minor pathway but one of great importance, because it is the activating pathway in terms of antiesterase activity (Reiner, 19811. Once dichlorvos has been formed, the dichlorovinyl moiety may be removed by esterases, resulting in a vinyl alcohol that tautomerizes to dichloroacetaldehyde. Demethyl- ation of dichlorvos also occurs, and the demethylated phosphate is apparently unstable, fragmenting into several products including di- chloroacetaldehyde (Dedek, 19811. Nordgren et al. (1981), on the basis of pharmacokinetic models of data from human ingestion, suggested that a majority of the plasma and erythrocyte trichlorfon is eliminated through the formation of dichlorvos. The authors stated that in this manner tri- chlorfon becomes a "slow release formulation" of dichlorvos and results in prolonged inhibition of esterases relative to the duration of action of dichlorvos (see below). It may also have an impact on the safety of the compound in that dichlorvos may be a deoxyribonucleic acid (DNA) alkylating compound, and dichloroacetate and dichloroacetaldehyde are known mutagens. Radiotracer studies indicate that lungs and urine are the major routes of excretion of trichlorfon in rats. Hassan et al. (1965) reported finding 28% of total radioactivity as SCOT and 32% of total radioactivity in the urine within 24 hours after a dose of ~4CH3O-labeled trichlorfon. When 32P-labeled compound was used, 75% to 85% of total radioactivity was recovered in the urine in 48 hours. They postulated that the urinary com- pounds were mainly di- and monomethyl phosphates and that the methoxy cleavage products were further metabolized to formate (a minor urinary metabolite) and from that compound to carbon dioxide (Hassan and Zayed, 1965). HEALTH ASPECTS Observations in Humans Trichlorfon has been used successfully as an antischistosomal agent in many humans. Oral therapeutic doses have ranged from 5 to 20 mg/kg and were formerly given for up to 23 consecutive days (see review by Holmstedt et al., 19781. The more recent dose regimen has been three doses separated by 2 to 4 weeks (Jewsbury et al., 1977; Reddy et al;, 19751. These therapeutic doses inhibit serum and red blood cell (RBC)

Toxicity of Selected Contaminants 399 cholinesterases to a great extent. Forsyth and Rashid (1967) reported approximately 90% inhibition of serum cholinesterase in three male pa- tients within 1 hour after ingesting 10-mg/kg doses of trichlorfon. Inhi- bition reversed slowly and was not complete (10% to 30% inhibition remained) after 2 weeks. Hanna et al. (1966) reported comparable inhi- bition 30 minutes after dosing 10 patients with 5 mg/kg. The purity of trichlorfon used in these studies was unreported. Nordgren et al. (1981) gave 98% pure trichlorfon to seven schistosomi- asis patients. The researchers observed the rapid appearance of both the parent compound and its putative active metabolite dichlorvos (in quan- tities some 100-fold less than the parent compound) in blood after a single oral dose between 7.5 and 10 mg/kg. Plasma cholinesterase fell to near zero activity within 15 minutes. Recovery rate was reported to be slow. RBC cholinesterase was less inhibited than serum esterase, but recovery of activity was slower than that recorded for serum esterase. The authors stated that a second dose of 10 mg/kg after 14 days produced more RBC esterase inhibition than the original treatment. Many of the reports on patients receiving trichlorfon therapy noted the presence of mild clinical signs of acetylcholinesterase inhibition. Gas- trointestinal disturbances appear to have been the most prevalent effect. Hanna et al. (1966) noted decreased sperm counts and sperm motility in 10 men treated with metrifonate (5 mg/kg for 12 consecutive days) for urinary schistosomiasis.This effect was also noted by Wenger (WHO, 1972). Certain organophosphorus esters produce axonal degeneration in pe- ripheral nerves and long axon tracts of the spinal cord—a neuromyelop- athy. Studies reviewed by Johnson (1981) indicate that humans ingesting high doses of trichlorfon have developed neurological dysfunction that appears to be organophosphorus neuropathy. The majority of these cases were reported in the Soviet literature, and Johnson questioned that im- purities with greater toxicity might have contaminated the trichlorfon. It is well known that the higher chain length analogs of dichlorvos, partic- ularly di-n-butyl- and di-n-pentyldichlorovinyl phosphate, are extremely neurotoxic and have little effect on acetylcholinesterase (Albert and Stearns, 19741. Nonetheless, in at least one case (an attempted suicide), a syndrome identical to organophosphorus neuropathy was produced by ingestion of trichlorfon that was shown to have no toxic impurities (Hierons and John- son, 19781. The dose that produced neuromyelopathy in this and other cases (e.g., Senanayake and Johnson, 1982) also caused severe clinical signs of acetylcholinesterase inhibition. Van Bao et al. (1974) studied five people exposed to high lev~els of Ditriphon-SO a Hungarian insecticide preparation containing trichlorfon as the active ingredient and other unknown constituents. They found an

400 DRINKING WATER AND HEALTH excess incidence of short-lived chromosome breaks and exchange figures relative to karyotypes of 15 healthy volunteers matched for age only. A significant increase in stable chromosome alterations was also found among the people exposed to Ditriphon 50. Observations in Other Species Acute Elects In a study with trichlorfon, Edson and Noakes (1960) reported an acute oral LDso at 649 mg/kg bw in male Wistar rats and a higher value in females. For dermal exposures, lethality (LDso) was less than 2,800 mg/kg. The authors presumed that the cause of death was related to the inhibition of acetylcholinesterase. A review by Holmstedt et al. (1978) indicated that there were comparable or lower LDsoS in rats, mice, guinea pigs, and dogs given presumably pure trichlorfon by a variety of parenteral routes (lowest value, 150 mg/kg from intravenous doses in dogs; highest value, 650 mg/kg from intraperitoneal exposures of mice). DuBois and Cotter (1955) administered purified trichlorfon intraperito- neally to female rats and observed approximately 50% inhibition of brain acetylcholinesterase within 15 minutes after a 25-mg/kg dose. (Serum cholinesterase was 64% inhibited at this point.) Normal acetylcholines- terase activity was noted 1 hour after the dose. Greater inhibition and slower recovery occurred at higher doses. Edson and Noakes (1960) ob- served only 30% inhibition of RBC and serum esterases after an acute 500-mg/kg dose of a 50% trichlorfon solution and reported recoveries comparable to those noted by DuBois and Cotter (19551. The disparity between these findings and those reported for humans is notable: humans appear more sensitive to esterase inhibition, and the recovery of enzyme activity is slower. Trichlorfon may potentiate the toxicity of other organophosphorus in- secticides. The effect is presumably due to inhibition of enzymes that detoxify certain organophosphorus compounds (e.g., malathion) and is transient (DuBois, 19581. Subchronic E~ects DuBois and Cotter (1955) maintained a steady- state 25% to 50% inhibition of rat brain acetylcholinesterase for 60 days with daily intraperitoneal 50-mg/kg doses of purified trichlorfon. At this dose level there were no deaths. However, 100 mg/kg caused progressive inhibition of acetylcholinesterase and death in two and five animals. All five rats died at doses of 150 mg/kg. In a 1955 study, Doull and DuBois (J. Doull, University of Kansas Medical Center, Kansas City, Kans., personal communication, 1985; WHO, 1972) reporte.d cholinesterase depression in rats (13 females and 13 males per group) fed 300-ppm dietary concentrations of trichlorfon for 16 weeks but no effect at 100 ppm,

Toxicity of Selected Contaminants 401 whereas Edson and Noakes (1960) observed no effects in male rats (10 per group) given up to 125-ppm concentrations of trichlorfon (11.3 ma/ kg/day) in the diet for 16 weeks. No changes in growth, food consumption, or gross appearance of tissues were observed at any dose in these two studies. WHO (1972) cited two unpublished studies by Doull. In a 1962 study, cholinesterase was depressed in dogs (two males and two females per group) fed 500-ppm but not 250-ppm dietary concentrations of trichlorfon for 1 year (Doull, personal communication, 19851. In a 1958 follow-up, Doull and Vaughn fed 20-, 100-, 300-, and 500-ppm concentrations of trichlorfon to dogs (one female and one male per group) for 12 weeks followed by 4 weeks on unadulterated feed. At the end of this study, there was significant depression of serum and RBC cholinesterase in the 300- ppm (7.5 mg/kg/day) group, but not in the 100-ppm (2.5 mg/kg/day) group (Doull, personal communication, 19851. These data are comparable to those of Williams et al. (1959), who reported that plasma and RBC cholinesterase were inhibited in dogs (one male and one female per group) given trichlorfon at 500-ppm (12.5 mg/kg/day) in the feed for 12 weeks but that no effects were noted in animals given 200 ppm (5 mg/kg/day) for the same period. In 1970 Loser (D. Lamb, Mobay Chemical Corpo- ration, Stilwell, Kans., personal communication, 1985; WHO, 1972) per- formed a 4-year feeding study in dogs (four males and four females per group), in which plasma and RBC cholinesterase were inhibited in groups fed 200 ppm (5 mg/kg/day) but not in groups given 50 ppm (1.25 ma/ kg/day). Males in a group fed 800 ppm (20 mg/kg/day) had enlarged spleens and reduced adrenal size. (These data are summarized in Table 9-24.) In 1981 Coulston studied the effects of subacute trichlorfon exposures in rhesus monkeys. Erythrocyte cholinesterase was reduced in monkeys given 1-mg/kg oral doses of trichlorfon for 4 years and "possibly" reduced at a dose of 0.2 mg/kg. No alteration of body weight or "clinical chem- istry" had been observed, although diarrhea was reported to occur at 5- mg/kg doses (F. Coulston, White Sands Research Center, Rensselaer, N.Y., personal communication, 19851. Chronic EJ5fects Most chronic studies on trichlorfon are unpublished reports that were summar~zed by WHO (1972) or by Machemer (19811. A synopsis of those data is given in this section. In 1962 Doull and colleagues fed Sprague-Dawley rats (25 of each sex) diets containing 0- to 1,000-ppm concentrations of technical-grade tri- chlorfon (Doull, personal communication, 1985; WHO, 19724. The dosing was intended to last for 24 months, but it was shortened to 17 months in males (due to mortality). Males in the 1,000-ppm (100 mg/kg/day) group

402 DRINKING WATER AND HEALTH TABLE 9-24 Effect of Subacute Administration of Tr~chlorfon on Cholinesterase Activities Level Producing Test Site of NOELa Inhibition Duration Animal Inhibition (mg/kg/day) (mg/kg/day) of Study Reference Rat Brain 50 NRb DuBois and Cotter, 1955 Rat Unknown 10.0 30 16 weeks Doull, personal communication, 1985C Rat RBC,4 plasma, 11.3 — NR Edson and Noakes, brain 1960 Dog Unknown 6.25 12.5 1 year Doull, personal communication, 1985C Dog RBC, serum 2.5 7.5 12 weeks Doull, personal communication, 1985C Dog RBC, plasma 5.0 12.5 12 weeks Williams et al., 1959 Dog RBC, plasma 1.25 5.0 4 years Lamb, personal communication, 1985C No-observed-effect level. bNR = not reported. CSee summary in WHO, 1972. Fred blood cell. failed to gain weight equivalent to other groups, and life span was short- ened in both sexes at the high dose. Serum cholinesterase, but not brain, submaxillary gland, or RBC cholinesterase, was inhibited by 25% in the group fed 500 ppm (50 mg/kg/day) but not in animals fed 250 ppm (25 mg/kg/day). Vascular lesions, including fibrous changes and necrotizing inflammation, were reported, but the dosing group in which they occurred was not identified. Histopathology was done on only five rats of each sex from each dose group. A similar study was undertaken by Doull and coworkers again in 1965 (Doull, personal communication, 1985; WHO 19721. They gave Sprague- Dawley rats (25 males and 50 females per group) 100-, 200-, or 400-ppm concentrations of technical-grade trichlorfon in feed. The study was in- tended to last 18 months, but was apparently curtailed at about 70 weeks due to mortality in both the treated and control groups. There was a sex difference in cholinesterase inhibition: slight depression of RBC cholin- esterase occurred in males at 200 ppm but not at 100 ppm (10 mglkgl day), whereas there was "very slightly" depressed R13C cholinesterase in females fed 100 ppm. Spleen and liver weights were reported to be reduced in the 400-ppm group (40 mg/kg/day), and four females in the

Toxicity of Selected Contaminants 403 high-dose group had pulmonary changes. (Other findings of this study are discussed in the pertinent sections below.) Doull's findings contrast with those of Lorke and Loser in 1966 (Lamb, personal communication, 1985; WHO, 1972) and of Grundman and Hobik in 1966 (Lamb, personal communication, 1985; WHO, 19721. These researchers fed Long-Evans rats (50 males and 50 females per treated group and 100 of each sex for controls) 50-, 250-, 500-, and 1,000-ppm dietary concentrations of technical-grade trichlorfon for 2 years. Cholin- esterase inhibition was seen only in the high-dose group. Coulston indi- cated that CD1 mice (60 females and 60 males per group) given trichlorfon for 90 weeks were unaffected by 100 ppm in the diet, but that 300 ppm and 1,000 ppm (45 and 150 mg/kg/day) decreased weight gain and cho- linesterase in both sexes (F. Coulston, White Sands Research Center, Rensselaer, N.Y., personal communication, 19851. Neurotoxicity Johnson (1970) and Olajos et al. (1979) produced clin- ical signs of neuropathy in adult laying hens by giving them 200 mg/kg trichlorfon followed 3 days later by a dose of 100 mg/kg (both doses subcutaneous). Both investigators had to protect the animals from cholin- ergic toxicity with atropine. (Johnson also used physostigmine and an oxime deactivator.) Mutagenicity The genetic toxicity of trichlorfon has been reviewed by IARC (1983, pp. 207-231) and by Moutschen-Dahmen et al. (19811. A disturbing aspect about the information relating to the genetic toxicity of trichlorfon is the number of contradictory reports regarding its activity in different assays. For most compounds, a profile of genetic activity for specific types of end points becomes clear as more tests are performed. The occasional contradictory result can usually be dismissed by examining differences in protocol or by demonstrating inconsistency with other re- ports. In the case of trichlorfon, opposite results are often reported with no clear indications as to which is correct. This may be related to its instability, to contaminants, or to different formulations of pesticides tested. Trichlorfon apparently has the ability to react directly with DNA in vitro (Rosenkranz and Rosenkranz, 19721. This property has been con- h~rmed in mice whose tissues were examined for 7-methylguanine follow- ing exposure to trichlorfon (Dedek, 1981; Dedek et al., 1975, 19761. The methylating capacity of trichlorfon was found to be 10-fold less than that of dichlorvos and 1,000-fold less than that of dimethyl sulfate. The binding in the testicles was approximately an order of magnitude lower than the binding in liver, kidneys, or lung. In similar studies (Segerback and Ehrenberg, 1981), the genetic risk to patients treated for schistosomiasis

404 DRINKING WATER AND HEALTH with a therapeutic 15-mg/kg dose of trichlorfon was estimated to be on the same order of magnitude as the risk from 100 mrad of gamma radiation. This is approximately equal to the annual background dose and was char- acterized as a "low or very low" risk. Both positive and negative results have been obtained in bacterial mu- tagenesis assays with trichlorfon. When activity is reported, it is usually in base-pair substitution strains without the need for exogenous metabolic activation, consistent with direct aLkylating activity (IARC, 1983, pp. 207-2311. There was some indication that mutagenic activity increased during storage, suggesting the production of a mutagenic breakdown prod- uct such as dichlorvos. There is one report that trichlorfon induces mutations in yeast (Gilot- Delhalle et al., 1983), and one report that it does not (Morpurgo et al., 19771. IARC (1983, pp. 207-231) reported mild activity inducing mitotic crossing over in yeast, without the need for exogenous metabolic acti- vation. Trichlorfon was negative in Drosophila mutagenesis assays. Be- cause of high toxicity, however, only low doses could be examined (IARC, 1983, pp. 207-231~. Results from tests for mutagenesis and cytogenetic damage in cultured mammalian cells are generally positive for trichlorfon (IARC, 1983, pp. 207-2311. Results from cytogenetic assays in mice with trichlorfon are, once again, mixed (IARC, 1983, pp. 207-2311: what appear to be well-conducted studies give negative results in both bone-marrow and spermatogonial chromosomes (Moutschen-Dahmen et al., 19811. Chro- mosome analysis was done on blood samples from several human subjects following self-intoxication by large doses of trichlorfon. Chromosome damage was approximately eightfold greater than that of controls soon after exposure, but returned to near control levels by 6 months (van Bao et al., 19741. The small numbers of subjects and the lack of appropriate controls prevent dete~ination of a causal effect in these cases (IARC, 1983, pp. 207-2311. Other reports of chromatic aberrations, such as sister- chromatid exchange (SCE), have been reported in factory workers exposed to trichlorfon. However, elevations in this end point were also seen in workers from the same plant not exposed to trichlorfon, suggesting air- borne exposure to other chemicals as well (IARC, 1983, pp. 207-2311. Reports of mutagenicity, along with the observation that trichlorfon can affect sperm morphology (Wyrobek and Bruce, 1975), indicate potential germ cell mutagenic activity for trichlorfon. Both positive and negative results have been reported with trichlorfon in the dominant lethal assay in mice (IARC, 1983, pp. 207-2311. In what appear to be well-conducted dominant lethal studies, no statistically significant activity of trichlorfon (Moutschen-Dahmen et al., 1981) or dichlorvos (Epstein et al., 1972) was observed.

Toxicity of Selected Contaminants 405 Information generated thus far suggests that trichlorfon is a weak, direct- acting mutagen. The assays may be influenced by cytotoxic activity and the presence of breakdown products or contaminants. The data are so mixed that it is difficult to conduct a risk assessment based on mutagenic activity at this time. Future research should include examination of the dose-response rela- tionship by studying the induction of cytogenetic damage in blood from patients before and after therapeutic exposure to trichlorfon. The potential germ cell mutagenicity of trichlorfon should be reexamined with an em- phasis on altered sperm morphology in patients before and after trichlorfon therapy for parasitic infestations. Carcinogenicity Bioassays of trichlorfon for carcinogenicity have pro- duced mixed results, and some studies have been criticized (Machemer, 19811. Furthermore, details of the older studies have not been reported, but have only been summarized in reviews (IARC, 1983, pp. 207-231; Machemer, 1981; WHO, 19721. IARC (1983) cited a carcinogenesis bioassay in progress, and the National Toxicology Program (NTP, 1984b) reported a carcinogenesis bioassay being conducted by the Gulf South Research Institute under contract to the National Institute of Environmental Health Sciences. NTP (1984a) reported a slight increase in mammary tumor incidence and earlier onset in female rats given 250-, 500-, and 1,000-ppm concentrations of trichlorfon in the diet for 24 months. No statistical analysis was reported by these investigators. Five animals from each dose group were examined microscopically. Mammary tumors of three different types were seen in three of the five 1,000-ppm animals; androblastoma was found in two of the five. Three animals from the 500-ppm group and two from the 250- ppm group had developed mammary tumors. In 1965, Doull and coworkers reported mammary tumors in 15%, 11%, and 8% of female rats (50 per group) fed trichlorfon for 18 months at a level of 400, 200, and 0 ppm, respectively (Doull, personal communication, 1985; WHO, 19721. Mi- croscopic examination of the tumors indicated that they were benign. In 1966 Lorke and Loser (Lamb, personal communication, 1985; WHO, 1972) found no increased incidence in malignant tumors in Long-Evans rats given up to 1,000-ppm concentrations of trichlorfon in feed for 2 years. Gibel et al. (1971, 1973) gavaged 40 Wistar rats (unspecified sex) with 15-mg/kg doses of trichlorfon twice weekly for life. The average life span for treated animals was shortened, and 7 of 28 treated animals had malignant tumors (of unspecified type) as compared to no tumors in con- trols. Benign tumors were found in 19 of 28 rats and in 3 of 36 controls. Teichmann and Hauschild (1978) gave AB/Jena mice (30 males and 28 females) intraperitoneal 30-mg/kg doses of trichlorfon twice weekly for

406 DRINKING WATER AND HEALTH 75 weeks and found no significant difference in tumor incidence between treated and control animals. The same investigators (Teichmann et al., 1978) gave albino rats (30 males and 35 females) 22 mg/kg intraperito- neally twice weekly for 90 weeks and reported no increased tumor inci- dence over controls (25 males and 26 females). Teichmann and Schmidt (1978) found no increased tumor incidence in hamsters (23 male and 25 female treated animals; 22 male and 23 female controls) given 20-mg/kg doses of trichlorfon intraperitoneally twice weekly for 90 weeks. The possible mechanisms of toxicity or interaction responsible for the majority of the effects described above are not known. In contrast, the mechanism by which acute toxicity proceeds, i.e., by inhibition of ace- tylcholinesterase, is understood at the molecular level. Inhibition of ace- tylcholinesterase by phosphorylation of its active site and the sequelae of acetylcholine accumulation are reviewed in Chapter 4. This action is probably a property of dichlorvos, a metabolite of trichlorfon (Reiner, 1981~. A comparable reaction probably occurs between dichlorvos and other esterases such as serum cholinesterase and some carboxylesterases. Developmental Effects Staples and Goulding (1979) reported terato- genic effects (edema, herniation of the brain, hydrocephaly, micrognathia, cleft palate, shortened radius or ulna, hypophalangism, and syndactyly) in rats and hamsters given trichlorfon. Malformations were seen in CD rats given 480 mg/kg/day (in three divided doses) on days 6 through 15 of gestation but not if doses were given only on day 8 or 10. Teratogenicity in hamsters was produced by doses of 400 mg/kg/day (in divided doses) on days 7 through 11 of gestation. At 300 mg/kg/day, fetus weight was reduced in hamsters; no effect was seen at 200/mg/kg/day. Doses of 400 mg/kg/day given on day 8 of hamster gestation were embryotoxic but not teratogenic. Increased incidence of cleft palate only was induced in CD1 mice given 600 mg/kg/day on days 10 through 14 or 12 through 14 of gestation. The doses that were teratogenic in hamsters and rats were higher than the dose required to produce cholinergic signs in the dams. The incidence of teratogenic effect was still significant when compared with food-restricted controls in which weight gain was depressed as much as in treated dams. Treated mice did not develop cholinergic signs or terata relative to the more sensitive species (rat and hamster). CD1 mice had mixed response to 200, 300, and 400 mg/kg/day (Courtney and Andrews, 19804. Fetuses had extra ribs at the low dose, were normal at 300 ma/ kg/day, and had fewer ribs at the high dose. Reproductive EJ:fects WHO (1972) reported that trichlorfon adminis- tered to rats at 1,000 and 3,000 ppm (100 and 300 mg/kg/day) had effects on reproduction, but details were sparse. The effect of trichlorfon on

Toxicity of Selected Contaminants 407 reproductive organs has been studied by two investigators with mixed results. Female rats were given trichlorfon at 500 or 1,000 ppm (50 or 100 mg/kg/day) in feed for 24 months (Doull, personal communication, 1985; Lamb, personal communication, 1985; WHO, 1972), and five an- imals per group were examined. The investigators found that the rats were without primitive ova or primary follicles. The same finding was reported in one of five females from the croup given 250 ppm (25 mg/kg/day). Two of five rats at 1,000 ppm (100 mg/kg/day) had tubular androblas- tomas. Three of five males examined from the 1,000-ppm group had focal aspermatogenesis. Cystic ovaries were observed in 40%, 33%, 14%, and 8% of rats given 400-, 200-, 100-, or O-ppm concentrations of trichlorfon in the diet for 70 weeks, respectively (Doull, personal communication, 1985; Lamb, personal communication, 1985; WHO, 1972~. Results of statistical analysis were not reported. Microscopic examination of tissue from four of five animals from the 400-ppm (40-mg/kg/day) group revealed the absence of follicles and ova. No change in gonads was found in Long- Evans rats given 50- to 1,000-ppm concentrations of trichlorfon in the diet for 24 months (Doull, personal communication, 1985; Lamb, personal communication, 1985; WHO, 19721. Dogs given trichlorfon at 1,000 ppm (25 mg/kg/day) in the diet for 1 year were reported to have a decrease in spermatogenesis (Doull, personal communication, 1985; Lamb, personal communication, 1985; WHO, 1972~. The investigators also reported de- creased gonad weight in male dogs fed trichlorfon at 800 ppm (20 ma/ kg/day) and females given 3,200 ppm (80 mg/kg/day) in the diet for 4 years. There were, however, no morphological changes in these organs. CONCLUSIONS AND RECOMMENDATIONS The human genotoxic risk from trichlorfon exposure is not at all clear from the reported studies on the carcinogenic, teratogenic, or reproductive effects of the compound in humans or animals. Large variability exists in the observations, quality of the studies, detail of the reports, and even the purity of the test compound. Multiple reports that trichlorfon or some of its metabolites or breakdown products are mutagenic suggest that at least the potential for genotoxicity exists. The committee therefore strongly recommends that a well-conducted carcinogenesis bioassay of trichlorfon be performed as well as a meticulous assessment of its teratogenicity and reproductive toxicity. Trichlorfon probably produces neurodegeneration (organophosphorus neuropathy), but the risk of this effect is small relative to other actions of the compound. The unequivocal effect that occurs at the lowest dose is inhibition of cholinesterases. That the inhibition of acetylcholinesterase by trichlorfon (or its metabolites) can produce a toxic syndrome is un-

408 OR ~ N. K! NG WATER AN D H EALTH deniable. The significance of inhibition of other esterases as a toxic effect in and of itself is yet to be determined. In a study with monkeys, Machemer (1981) reported the lowest dose in which an effect on cholinesterase was observed (0.2 mg/kg/day for 4 years). Extrapolation to humans from data on nonhuman primates would be desirable; however, because the study has only been summarized, there is not enough detail to evaluate for this purpose. Among studies reported in more detail, the lowest concentration at which an action of trichlorfon on cholinesterases has been reported is 200 ppm (5 mg/kg/day) in the feed of dogs (eight per group) for 4 years (Doull, personal communication, 1985; Lamb, personal communication, 1985; WHO, 19721. The highest no-effect concentration in this study was 50 ppm (1 .25 mg/kg). Again, this report has only been summarized, although in a more complete manner. It would be better if an actual level of inhibition had been described in the report, but it is assumed that fairly significant inhibition of serum cholinesterase was encountered. This opin- ion is based on the observation that in most studies serum cholinesterase tended to be more sensitive to trichlorfon than was RBC cholinesterase and that the latter enzyme was inhibited as well (Doull, personal com- munication, 1985; Lamb, personal communication, 1985; WHO, 19721. In determining a safety factor to be applied to the dose level, it is worth recalling that interspecies differences in trichlorfon effects may have ex- isted both within and between studies. More importantly, comparison of rat and human response indicates that humans are more sensitive than rodents in terms of magnitude and duration of cholinesterase inhibition. The committee recommends a conservative safety factor of 100. Assuming that a 70-kg human consumes 2 liters of water daily and that 20% of the intake of trichlorfon is derived from water, one may calculate the suggested acceptable daily intake (ADI) as: 1.25 mg/kg low/day x 70 kg x 0.2 _ 0.088 mg/liter, or 88 ~g/liter. 2 liters x 100 Hi. An ADI may also be estimated for a 10-kg child on the assumption that ~ liter of water is consumed daily and that 20% of the trichlorfon intake is derived from water: 1.25 mg/kg/day x 10 kg x 0.2 1 liter x 100 _ 0.025 mg/liter, or 26 ~g/liter. The committee recommends that a new assay for carcinogenicity of trichlorfon be conducted following currrent protocols and methods. In addition, monitoring of workers exposed to trichlorfon for genetic effects such as SCE should continue.

Toxicity of Selected Contaminants 409 CONCLUSIONS AND RECOMMENDATIONS Table 9-25 lists the substances reviewed in this volume and presents either an ADI, a SNARL, or cancer risk estimate where data were sufficient for calculation. The statistical methodology for the risk estimation is de- scribed in Chapter 8. Further information on methodology can also be found in Volumes 1 and 3 of Drinking Water and Health (NRC, 1977, 1980~. When reviewing this table, the reader should refer back to the discussion on individual compounds for specific details. The committee recommends that the following studies be conducted: Acrylamide: · Chronic studies to evaluate the neuropathic effects of acrylamide. Aldicarb: · Research on the unconfirmed effects of low-level inhibition of ace- tylcholinesterase and other esterases. Diallate: · A carcinogenicity study using an improved protocol with more than one dose level. · Studies of metabolism and developmental effects. D ibromochloropropane: · Oral feeding studies to determine clear dose-response relationships for reproductive effects. Di(2-ethylhexylJ Phthalate (DEHP): · Investigations of the role of DEHP-induced hepatocellular changes, such as peroxisomal and cellular proliferation, in carcinogenesis. · Examination of the structure-activity relationships of DEHP and sim- ilar surfactants with cell membrane-altering properties in cell transfor- mation bioassays and correlation of those relationships with aneuploidy induction as well as carcinogenicity. 1, 2 -Dichloropropane, I, 2, 3-Trichloropropane, and 1, 3-Dichloropro- vene: · Additional studies on the toxicity of the chloropropanes and chlo- ropropenes, focusing on pure substances rather than on mixtures. Ethylene Dibromide: · A carcinogenicity bioassay to demonstrate a reliable dose response from oral exposure. · Additional studies on reproductive toxicity.

410 a: Ct Cal ¢ by V, _' Cal a: o is: Cal 1 Cal a; ¢ o Ad A_ Cal V) C) - o _` Cal Cal .— 3 .O Cal ¢ ~ JO V o o Cd CQ U) ~ Cal S-1 Cal V o v ~ ·= & C'D ._ A_ ~4 C ·- Cal ~ ~ _1 5 ~ a ·— Cal hi O ~ U: ._ EM in: ._ lo: o Cal Cal o & o v \~~ Sc~ \ 1 1 1 1 1 o o o o o ~ ~ ~ - ~ ~ x x x x x n } —~ O~ · ~ ~o ~ s~ . - . - - . . o ~ c) c) . - . - - ~ oo o o o o o . . o o ;^ e 'e ;^ e e~ b4 - - ~ ~ D s: ~ oc ~ o o ~ . . o o ) c~ ca S~ ~ 'e ' c) e ~ ~ ~ D _ j _ _ · 2 ~ o ° e ~ ~ & O Ct O 6 :=, <~e -= 2 a° ~ a -u '

411 x x ~ c-) ;^ ~ ~ ~ - c) ~ ~ o D ._ At Cal c C ~ 2 o ~ ~ _ ~ Adz US to Ct ~ 1 .c o o C) AD _ i., ~ ~ Ct c ¢) ~ _ _ ._ ._ _ _ oo C~ ·— C) b4 — . ~ . _ o.e e~ oo ~ oo ·_ >~ >~ >, C~ .` _ ~ ~ C~ C C ~ .,°. O oD O4 e~4 C~ D · ~ ~ _ O ~ . _ O .= C) ~ ~ V~ ~E-Q o o .O C) ~ ~ ~ ._ C) ~ ._ C~ ._ ._ - ~: o - au C~ C~ ._ Ct o C~ o o C~ .= O C) s- O ~ ~ ~ ,> ) · e ct Ce 'e >~ o o ) ', ~ o o c~ "= " ~ ~ c z <~ t~ c~ o o c) c) C~ ~ ~ ~ J O - oo

4~2 DRINKING WATER AND HEALTH Nitrofen: · Additional studies on mechanisms of teratogenicity and carcinoge- nicity. Pentachlorophenol: · Additional carcinogenicity bioassays. · Studies to resolve uncertainties concerning the effects of contami- nants, e.g., chlorinated dioxins and dibenzofurans. Sulfallate: · Additional carcinogenicity studies with improved protocols. · Additional studies on metabolism and developmental effects. Trichlorfon: · A well-conducted carcinogenesis bioassay. · Assessments of teratogenicity and reproductive toxicity. REFERENCES Abe, S., and M. Sasaki. 1977. Chromosome aberrations and sister chromatic exchanges in Chinese hamster cells exposed to various chemicals. J. Natl. Cancer Inst. 58:1635- 1641. Agarwal, D. K., S. Eustis, J. C. Lamb, IV, J. R. Reel, and W. M. Kluwe. 1985. Correlation of dose-dependent functional pathophysiological changes induced by di-(2-ethylhexyl)- phthalate (DEHP). (Abstract 467.) Toxicologist 5: 117. Agrawal, A. K., P. K. Seth, R. E. Squibb, H. A. Tilson, L. L. Uphouse, and S. C. Bondy. 1981. Neurotransmitter receptors in brain regions of acrylamide-treated rats. I: Effects of a single exposure to acrylamide. Pharmacol. Biochem. Behav. 14:527-531. Ahlborg, U. G., J.-E. Lindgren, and M. Mercier. 1974. Metabolism of pentachlorophenol. Arch. Toxicol. 32:271 -281. Albert, J. R., and S. M. Stearns. 1974. Delayed neurotoxic potential of a series of alkyl esters of 2,2-dichlorovinyl phosphoric acid in the chicken. (Abstract 156.) Toxicol. Appl. Pharmacol. 29:136. Albro, P. W., and B. Moore. 1974. Identification of the metabolites of simple phthalate diesters in rat urine. J. Chromatogr. 94:209-218. Albro, P. W., R. Thomas, and L. Fishbein. 1973. Metabolism of diethylhexyl phthalate by rats: Isolation and characterization of the urinary metabolites. J. Chromatogr. 76:321- 330. Albro, P. W., J. R. Hass, C. C. Peck, D. G. Odam, J. T. Corbett, F. J. Bailey, H. E. Blatt, and B. B. Barrett. 1981. Identification of the metabolites of di-(2-ethylhexyl) phthalate in urine from the African green monkey. Drug Metab. Dispos. 9:223-225. Albro, P. W., J. T. Corbett, J. L. Schroeder, S. Jordan, and H. B. Matthews. 1982. Pharmacokinetics, interactions with macromolecules and species differences in metab- olism of DEHP. Environ. Health Perspect. 45:19-25. Alumot (Olomucki), E., and Z. Harduf. 1971. Impaired uptake of labeled proteins by the ovarian follicles of hens treated with ethylene dibromide. Comp. Biochem. Physiol. 39B :61 -68.

Toxicity of Selected Contaminants 413 Alumot (Olomucki), E., E. Nachtomi, O. Kempenich-Pinto, E. Mandel, and M. Schindler. 1968. The effect of ethylene dibromide in feed on the growth, sexual development and fertility of chickens. Poult. Sci. 47:1979-1985. Ambrose, A. M., P. S. Larson, J. F. Borzelleca, R. B. Smith, Jr., and G. R. Hennigar, Jr. 1971.Toxicologic studies on 2,4-dichlorophenyl-p-nitrophenyl ether. Toxicol. Appl. Pharmacol. 19:263-275. Amer, S. M., and E. M. All. 1969. Cytological effects of pesticides. IV. Mitotic effects of some phenols. Cytologia 34:533-540. Amir, D., C. Esnault, J. C. Nicolle, and M. Courot. 1977. DNA and protein changes in the spermatozoa of bulls treated orally with ethylene dibromide. J. Reprod. Fertil. 51 :453- 456. Anonymous. 1977. PCP pesticides suspended in Michigan; dioxin found in cow livers. Pestic. Tox. Chem. News 5(16):30-32. Arrhenius, E., L. Renberg, L. Johansson, and M.-A. Zetterqvist. 1977. Disturbance of microsomal detoxication mechanisms in liver by chlorophenol pesticides. Chem.-Biol. Interact. 18:35-46. Arsenault, R. D. 1976. Pentachlorophenol and contained chlorinated dibenzodioxins in the environment: A study of environmental fate, stability, and significance when used in wood preservation. Proc. Annul Meet. Am. Wood-Preserv. Assoc. 72:122-148. Arthur, B. W., and J. E. Casida. 1957. Metabolism and selectivity of O,O-dimethyl 2,2,2- trichloro- 1-hydroxyethyl phosphonate and its acetyl and vinyl derivatives. J. Agric. Food Chem. 5:186-192. Ashby, J., F. J. de Serres, M. Draper, M. Ishidate, Jr., B. H. Margolin, B. Matter, and M. D. Shelby. 1985. Overview and conclusions of the IPCS collaborative study on in vitro assay systems. Prog. Mutat. Res. 5:117-174. Auld, R. B., and S. F. Bedwell. 1967. Peripheral neuropathy with sympathetic overactivity from industrial contact with acrylamide. Can. Med. Assoc. J. 96:652-654. Baader, H. C. E. W., and H. J. Bauer. 1951. Industrial intoxication due to pentachloro- phenol. Ind. Med. Surg. 20:286-290. Babich, H., D. L. Davis, and G. Stotzky. 1981. Dibromochloropropane (DBCP): A review. Sci. Total Environ. 17:207-221. Barrett, J. C., and P. W. Lamb. 1985. Tests with the Syrian hamster embryo cell trans- formation assay. Prog. Mutat. Res. 5:623-628. Barrett, J. C., A. Wong, and J. A. McLachlan. 1981. Diethylstilbestrol induces neoplastic transformation without measurable gene mutation at two loci. Science 212:1402-1404. Barrett, J. C., T. W. Hesterberg, and D. G. Thomassen. 1984. Use of cell transformation systems for carcinogenicity testing and mechanistic studies of carcinogenesis. Pharmacol. Rev. 36(Suppl.):53S-70S. Begley, J., E. L. Reichert, M. N. Rashad, and H. W. Klemmer. 1977. Association between renal function tests and pentachlorophenol exposure. Clin. Toxicol. 11:97-106. Bergner, H., P. Constantinidis, and J. H. Martin. 1965. Industrial pentachlorophenol poisoning in Winnipeg. Can. Med. Assoc. J. 92:448-451. Bermudez, E., J. C. Mirsalis, and H. C. Eales. 1982. Detection of DNA damage in primary cultures of rat hepatocytes following in vivo and in vitro exposure to genotoxic agents. Environ. Mutagen. 4:667-679. Bevenue, A., J. Wilson, L. J. Casarett, and H. W. Klemmer. 1967. A survey of pen- tachlorophenol content in human urine. Bull. Environ. Contam. Toxicol. 2-319-332. Biava, C. G., E. A. Smuckler, and D. Whorton. 1978. The testicular morphology of individuals exposed to dibromochloropropane. Exp. Mol. Pathol. 29:448-458.

414 DRINKING WATER AND HEALTH Bieri, F., P. Bentley, F. Waechter, and W. Staubli. 1984. Use of primary cultures of adult rat hepatocytes to investigate mechanisms of action of nafenopin, a hepatocarcinogenic peroxisome proliferator. Carcinogenesis 5:1033-1039. Biles, R. W., T. H. Connor, N. M. Tneff, and M. S. Legator. 1978. The influence of contaminants on the mutagenic activity of dibromochloropropane (DBCP). J. Environ. Pathol. Toxicol. 2:301-312. Bodansky, A. 1937. Determination of serum inorganic phosphate and of serum phosphatase. Am. J. Clin. Pathol. 7(Suppl. I):51-59. Bondi, A., E. Olomucki, and M. Calderon. 1955. Problems connected with ethylene dibromide fumigation of cereals. II. Feeding experiments with laying hens. J. Sci. Food Agric. 6:600-602. Botti, B., M. T. Moslen, N. M. Tneff, and E. S. Reynolds. 1982. Transient decrease of liver cytosolic glutathione S-transferase activities in rats given 1,2-dibromoethane or CC14. Chem.-Biol. Interact. 42:259-270. Boutwell, R. K., and D. K. Bosch. 1959. The tumor-promoting action of phenol and related compounds for mouse skin. Cancer Res. 19:413-424. Braun, W. H., J. D. Young, G. E. Blau, and P. J. Gehring. 1977. The pharmacokinetics and metabolism of pentachlorophenol in rats. Toxicol. Appl. Pharmacol. 41:395-406. Braun, W. H., G. E. Blau, and M. B. Chenoweth. 1978. The metabolism/pharmacokinetics of pentachlorophenol in man, and a comparison with the rat and monkey model. (Abstract 135.) Toxicol. Appl. Pharmacol. 45:278. BRL (Bionetics Research Laboratories). 1968a. Evaluation of Carcinogenic, Teratogenic, and Mutagenic Activities of Selected Pesticides and Industrial Chemicals. Volume I: Carcinogenic Study. Report No. NCI-DCCP-CG-1973-1-1. National Cancer Institute, Bethesda, Md. 401 pp. (Available from the National Technical Information Service, Springfield, Va., as Publication No. PB-223-159.) BRL (Bionetics Research Laboratories). 1968b. Evaluation of Carcinogenic, Teratogenic, and Mutagenic Activities of Selected Pesticides and Industrial Chemicals. Volume II: Teratogenic Study in Mice and Rats. Report No. NCI-DCCP-CG-1973-1-2. National Cancer Institute, Bethesda, Md. 150 pp. (Available from the National Technical Infor- mation Service, Springf~eld, Va., as Publication No. PB-223-160.) Broda, C., E. Nachtomi, and E. Alumot (Olomucki). 1976. Differences in liver morphology between rats and chicks treated with ethylene dibromide. Gen. Pharmacol. 7:345-348. Bull, R. J., M. Robinson, R. D. Laurie, G. D. Stoner, E. Greisiger, J. R. Meier, and J. Stober. i984. Carcinogenic effects of acrylamide in Sencar and A/J mice. Cancer Res. 44:107-111. Burek, J. D., R. R. Albee, J. E. Beyer, T. J. Bell, R. M. Carreon, D. C. Morden, C. E. Wade, E. A. Hermann, and S. J. Gorzinski. 1980. Subchronic toxicity of acrylamide administered to rats in the drinking water followed by up to 144 days of recovery. J. Environ. Pathol. Toxicol. 4(5):157-182. Burke Hurt, S. S., J. M. Smith, and A. W. Hayes. 1983. Nitrofen: A review and perspective. Toxicology 29:1-37. Buselmaier, W., G. Rohrborn, and P. Propping. 1972. Mutagenicity investigations with pesticides in the host-mediated assay and the dominant lethal test in mice. Biol. Zentralbl. 91:311-325. (In German; English summary) Butterworth, B. E., E. Bermudez, T. Smith-Oliver, L. Earle, R. Cattley, J. Martin, J. A. Popp, S. Strom, R. Jirtle, and G. Michalopoulos. 1984. Lack of genotoxic activity of di(2-ethylhexyl)phthalate (DEHP) in rat and human hepatocytes. Carcinogenesis 5:1329- 1335.

Toxicity of Selected Contam ~inants 415 Callaway, S., D. R. Davies, and J. P. Rutland. 1951. Blood cholinesterase levels and range of personal variation in a healthy adult population. Br. Med. J. 2:812-816. Cambon, C., C. Declume, and R. Derache. 1979. Effect of the insecticidal carbamate derivatives (carbofuran, pirimicarb, aldicarb) on the activity of acetylcholinesterase in tissues from pregnant rats and fetuses. Toxicol. Appl. Pharmacol. 49:203-208. Carere, A., and G. Morpurgo. 1981. Comparison of the mutagenic activity of pesticides in vitro in various short-term assays. Prog. Mutat. Res. 2:87-104. Carpenter, C. P., C. S. Well, and H. F. Smyth, Jr. 1953. Chronic oral toxicity of di(2- ethylhexyl) phthalate for rats, guinea pigs, and dogs. AMA Arch. Ind. Hyg. Occup. Med. 8:219-226. Casarett, L. J., A. Bevenue, W. L. Yauger, Jr., and S. A. Whalen. 1969. Observations on pentachlorophenol in human blood and urine. Am. Ind. Hyg. Assoc. J. 30:360-366. Cater, B. R., M. W. Cook, S. 13. Gangolli, and P. Grasso. 1977. Studies on dibutyl phthalate-induced testicular atrophy in the rat: Effect on zinc metabolism. Toxicol. Appl. Pharmacol. 41:609-618. Catilina, P. 1981. Atteintes sanguines par utilisations domestiques et professionnelles de produits xyloprotecteurs associant lindane et pentachlorophenol. Arch. Mall Prof. Med. Trav. Secur. Soc. 42:329-331. Catilina, P., A. Chamoux, M. J. Catilina, and J. Champeix. 1981. Contribution a lt etude des proprietes pathogenes de substances utilisees comme protecteurs des bois: Exemple du pentachlorophenol. Arch. Mall Prof. Med. Trav. Secur. Soc. 42:334-337. Cautreels, W., K. van Cauwenberghe, and L. A. Guzman. 1977. Comparison between the organic fraction of suspended matter at a background and an urban station. Sci. Total Environ. 8:79-88. Cerutti, P. A. 1985. Prooxidant states and tumor promotion. Science 227:375-381. Chapman, J. B., and P. Robson. 1965. Pentachlorophenol poisoning from bath-water. Lancet 1: 1266- 1267. Cherry, A. B., A. J. Gabaccia, and H. W. Senn. 1956. The assimilation behavior of certain toxic organic compounds in natural water. Sewage Ind. Wastes 28:1137-1146. Chiappino, G., and G. C. Secchi. 1968. Description of a case of acute poisoning from accidental ingestion of 1,2-dichloropropane, sold as trilene. Med. Lav. 59:334-341. (In Italian; English summary) Climie, I. J. G., D. H. Hutson, B. J. Morrison, and G. Stoydin. 1979. Glutathione conjugation in the detoxication of (Z)-1,3-dichloro-propene (a component of the nema- tocide D-D) in the rat. Xenobiotica 9:149-156. CMA (Chemical Manufacturers Association). 1982. Phthalate Esters Program Panel, Vol- untary Test Program, Health Effects Testing, Phase I: Validation Results. Vol. I. Chem- ical Manufacturers Association, Washington, D.C. [402 pp.] Cope, R. W., and R. R. Romine. 1973. TEMIK 10G Aldicarb Pesticide. Results of Aldicarb Ingestion and Exposure Studies with Humans and Results of Monitoring Human Exposure in Working Environments. Project Report No. 18269. Union Carbide Agricultural Prod- ucts Company, Inc., South Charleston, W. Va. [97 pp.] Costlow, R. D., and J. M. Manson. 1981. The heart and diaphragm: Target organs in the neonatal death induced by nitrofen (2,4-dichlorophenyl-p-nitrophenyl ether). Toxicology 20:209-227. Costlow, R. D., J. M. Hirsekorn, R. G. Stiratelli, G. P. O'Hara, D. L. Black, W. W. Kane, S. S. Burke, J. M. Smith, and A. W. Hayes. 1983. The effects on rat pups when nitrofen (4-(2,4-dichlorophenoxy) nitrobenzene) was applied dermally to the dam during organogenesis. Toxicology 28:37-50.

4~6 DRINKING WATER AND HEATH Courtney, K. D., and J. E. Andrews. 1980. Extra ribs indicate fetotoxicity and malde- velopment. (Abstract.) Teratology 21:35A. Croll, B. T., G. M. Arkell, and R. P. J. Hodge. 1974. Residues of acrylamide in water. Water Res. 8:989-993. Curto, K. A., and J. A. Thomas. 1982. Comparative effects of diethylhexyl phthalate or monoethylhexyl phthalate on male mouse and rat reproductive organs. Toxicol. Appl. Pharmacol. 62: 121-125. Daniel, J. W., and H. Bratt. 1974. The absorption, metabolism and tissue distribution of di(2-ethylhexyl)phthalate in rats. Toxicology 2:51-65. Davenport, J. G., D. F. Farrell, and S. M. Sumi. 1976. 'Giant axonal neuropathy' caused by industrial chemicals: Neurofilamentous axonal masses in man. Neurology 26:919- 923. Debets, F. M. H., J. J. T. W. A. Strick, and K. Olie. 1980. Effects of pentachlorophenol on rat liver changes induced by hexachlorobenzene, with special reference to prophyria, and alterations in mixed function oxygenates. Toxicology 15:181-195. Dedek, W. 1981. Guanine N7-alkylation in mice in To by metrifonate Discussion of possible genotoxic risk in mammals. Acta Pharmacol. Toxicol. 49(Suppl. V):40-50. Dedek, W., and K. Lohs. 1970. Zur alkylierenden Wirkung von Trichlorphon in Warm- blutern. II. Verteilung von '4C in Organen und Leberproteinen bei Ratten nach Appli- kation von i4C-Trichlorphon. Zeitschr. Naturforsch. 25b: 1110- 1113. Dedek, W., H. Koch, G. Uhlenhut, and F. Brose. 1969. Zur Kenntnis der Umsetzung von 3H-Trichlorphon zu DDVP. Zeitschr. Naturforsch. 24b:663-664. Dedek, W., K. Lohs, and W. Gibel. 1975. Studies on the alkylating properties of tri- chlorphone and dimethoate in mammals. Pp. 648-651 in Pesticides. Lectures Held at the IUPAC Third International Congress of Pesticide Chemistry, Helsinki. Georg Thieme, Stuttgart, Federal Republic of Germany. Dedek, W., K. Lohs, G. W. Fischer, and R. Schmidt. 1976. Alkylation of guanine in mice in vivo by organophosphorus insecticides. I. Trichlorphone and butonate. Pestic. Biochem. Physiol. 6:101-110. Deichmann, W., W. Machle, K. V. Kitzmiller, and G. Thomas. 1942. Acute and chronic effects of pentachlorophenol and sodium pentachlorophenate upon experimental animals. J. Pharmacol. Exp. Ther. 76:104-117. Deleu, R., and A. Copin. 1979. Extraction and gas-liquid chromatographic determination of the herbicides Nitrofen and Neburon: Application to natural waters. J. Chromatogr. 171 :263-268. (In French) de Lorenzo, F., S. degl'Innocenti, A. Ruocco, L. Silengo, and R. Cortese. 1977. Muta- genicity of pesticides containing 1,3-dichloropropene. Cancer Res. 37:1915-1917. de Lorenzo, F., N. Staiano, L. Silengo, and R. Cortese. 1978. Mutagenicity of diallate, sulfallate, and triallate and relationship between structure and mutagenic effects of car- bamates used widely in agriculture. Cancer Res. 38:13-15. DePass, L. R., E. J. Mirro, and F. R. Frank. 1982. Aldicarb Sulfone; Aldicarb Sulfoxide. Twenty-Nine-Day Water Inclusion Study in Rats. Project Report 45-18. Mellon Institute and Union Carbide Corporation, Bushy Run Research Center, Export, Pa. [42 pp.] Dietz, F. K., E. A. Hermann, and J. C. Ramsey. 1984. The pharmacokinetics of i4C-1,3- dichloropropene in rats and mice following oral administration. (Abstract 585.) Toxi- cologist 4:147. DiRenzo, A. B., A. J. Gandolfi, and I. G. Sipes. 1982. Microsomal bioactivation and covalent binding of aliphatic halides to DNA. Toxicol. Lett. 11:243-252. Doloshitskii, S. L. 1969. Hygiene standard for the chlorine-containing herbicide Avadex in water. Hyg. Sanit. 34(12):356-360.

Toxicity of Selected Contaminants 417 Domon, O. E., C. M. Teaf, B. Bhooshan, and J. B. Bishop. 1984. Assessment of fetal death rate among in utero progeny of B6C3 and CD-1 mice after subcutaneous injections of males with di(2-ethylhexyl)phthalate (DEHP). (Abstract.) Paper presented at the In- ternational Conference on Phthalic Acid Esters, August 6-7, 1984, University of Surrey, Guilford, Surrey, England. Doolittle, D. J., E. Bermudez, P. K. Working, and B. E. Butterworth. 1984. Measurement of genotoxic activity in multiple tissues following inhalation exposure to dimethylnitros- amine. Mutat. Res. 141:123-127. Dougherty, R. C. 1978. Human exposure to pentachlorophenol. Pp. 351-361 in K. R. Rao, ed. Pentachlorophenol: Chemistry, Pharmacology, and Environmental Toxicology. Plenum, New York. Doull, J., B. A. Bridges, R. Kroes, L. Golberg, I. C. Munro, O. E. Paynter, H. C. Pilot, R. Squire, G. M. Williams, and W. J. Darby. 1983. The Relevance of Mouse Liver Hepatoma to Human Carcinogenic Risk. A Report of the International Expert Advisory Committee to the Nutrition Foundation. The Nutrition Foundation, Inc., Washington, D.C. [46 pp.] Dow Chemical USA. 1984. Acrylamide: Environmental fate. Technical Report. Dow Chem- ical USA, Designed Products Department, Midland, Mich. 6 pp. Drew, R. T., J. M. Patel, and F.-N. Lin. 1978. Changes in serum enzymes in rats after inhalation of organic solvents singly and in combination. Toxicol. Appl. Pharmacol. 45:809-819. Dubelman, S., R. Lauer, D. D. Arras, and S. A. Adams. 1982. Operator exposure mea- surements during application of the herbicide diallate. J. Agric. Food Chem. 30:528- 532. DuBois, K. P. 1958. Potentiation of the toxicity of insecticidal organic phosphates. AMA Arch. Ind. Health 18:488-496. DuBois, K. P., and G. J. Cotter. 1955. Studies on the toxicity and mechanism of action of dipterex. AMA Arch. Ind. Health 11:53-60. Edson, E. F., and D. N. Noakes. 1960. The comparative toxicity of six organophosphorus insecticides in the rat. Toxicol. Appl. Pharmacol. 2:523-539. Edwards, K., H. Jackson, and A. R. Jones. 1970. Studies with alkylating esters. II. A chemical interpretation through metabolic studies of the antifertility effects of ethylene dimethanesulphonate and ethylene dibromide. Biochem. Pharmacol. 19:1783-1789. Edwards, P. M. 1975. The distribution and metabolism of acrylamide and its neurotoxic analogues in rats. Biochem. Pharmacol. 24:1277-1282. Edwards, P. M. 1976. The insensitivity of the developing rat foetus to the toxic effects of acrylamide . Chem. -Biol . Interact. 12: 13- 18. Egnatz, D. G., M. G. Ott, J. C. Townsend, R. D. Olson, and D. B. Johns. 1980. DBCP and testicular effects in chemical workers: An epidemiological survey in Midland, Mich- igan. J. Occup. Med. 22:727-732. Engelhardt, G., P. R. Wallnofer, and O. Hutzinger. 1975. The microbial metabolism of di-n-butyl phthalate and related dialkyl phthalates. Bull. Environ. Contam. Toxicol. 13:342-347. EPA (U.S. Environmental Protection Agency). 1977. Rebuttable presumption against reg- istration and continued registration of pesticide products containing diallate. Fed. Regist. 42:27669-27674. EPA (U.S. Environmental Protection Agency). 1978. Rebuttable presumption against reg- istration of pentachlorophenol (PCP) and its derivatives. Fed. Regist. 43:48443. EPA (U.S. Environmental Protection Agency). 1984a. 2,4-Dichlorophenyl-p-nitrophenyl ether; tolerances for residues. (40 CFR §180.223)

418 DRINKING WATER AND HEALTH EPA (U.S. Environmental Protection Agency). 1984b. Drinking Water Criteria Document for Pentachlorophenol. Draft prepared for the Office of Drinking Water by the Environ- mental Criteria and Assessment Office. U.S. Environmental Protection Agency, Cin- cinnati. [108 pp. ] EPA (U.S. Environmental Protection Agency). 1984c. Wood Preservative Pesticides: Creo- sote, Pentachlorophenol, Inorganic Arsenicals. Position Document 4. Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency, Washington, D.C. 344 PP Epstein, S. S., E. Arnold, J. Andrea, W. Bass, and Y. Bishop. 1972. Detection of chemical mutagens by the dominant lethal assay in the mouse. Toxicol. Appl. Pharmacol. 23:288- 325. Exon, J. H., and L. D. Koller. 1982. Effects of transplacental exposure to chlorinated phenols. Environ. Health Perspect. 46:137-140. Exon, J. H., and L. D. Koller. 1983a. Alteration of transplacental carcinogenesis by chlorinated phenols. Pp. 1177-1188 in R. L. Jolley, W. A. Brungs, J. A. Cotruvo, R. B. Cumming, J. S. Mattice, and V. A. Jacobs, eds. Water Chlorination: Environmental Impact and Health Effects. Vol. 4. Book 2: Environment, Health, and Risk. Ann Arbor Science, Ann Arbor, Mich. Exon, J. H., and L. D. Koller. 1983b. Effects of chlorinated phenols on immunity in rats. Int. J. Immunopharmacol. 5:131-136. Fahrig, R., C.-A. Nilsson, and C. Rappel 1978. Genetic activity of chlorophenols and chlorophenol impurities. Pp. 325-338 in K. R. Rao, ed. Pentachlorophenol: Chemistry, Pharmacology, and Environmental Toxicology. Plenum, New York. Fanini, D., M. S. Legator, and P. M. Adams. 1984. Effects of paternal ethylene dibromide exposure on Fat generation behavior in the rat. Mutat. Res. 139:133-138. Fleischer, M., R. Meiss, H. Robenek, H. Themann, and R. Eckard. 1980. Ultrastructural morphometric investigations on rat liver of young and adult rats after treatment with technical pentachlorophenol (PCP). Arch. Toxicol. 44:243-257. Forsyth, D. M., and C. Rashid. 1967. Treatment of urinary schistosomiasis with trichlo- rophone. Lancet 2:909-912. Foster, P. M. D., L. V. Thomas, M. W. Cook, and S. D. Gangolli. 1980. Study of the testicular effects and changes in zinc excretion produced by some n-alkyl phthalates in the rat. Toxicol. Appl. Pharmacol. 54:392-398. Foster, P. M. D., J. R. Foster, M. W. Cook, L. V. Thomas, and S. D. Gangolli. 1982. Changes in ultrastructure and cytochemical localization of zinc in rat testis following the administration of di-n-pentyl phthalate. Toxicol. Appl. Pharmacol. 63:120-132. Frazelle, J. H., D. J. Abernethy, and C. J. Boreiko. 1984. Enhanced sensitivity of the C3H/10T2 cell transformation system to alkylating and chemotherapeutic agents by treatment with 12-O-tetradecanoylphorbol-13-acetate. Environ. Mutagen. 6:81-89. Fullerton, P. M. 1969. Electrophysiological and histological observations on peripheral nerves in acrylamide poisoning in man. J. Neurol. Neurosurg. Psychiatry 32:186-192. Fullerton, P. M., and J. M. Barnes. 1966. Peripheral neuropathy in rats produced by acrylamide. Br. J. Ind. Med. 23:210-221. Gage, J. C. 1967. The signif~cance of blood cholinesterase activity measurements. Residue Rev. 18:159-173. Gangolli, S. D. 1982. Testicular effects of phthalate esters. Environ. Health Perspect. 45:77-84.

Toxicity of Selected Contaminants 419 Ganning, A. E., U. Brunk, and G. Dallner. 1984. Phthalate esters and heir effect on the liver. Hepatology 4:541-547. Garland, T. O., and M. W. H. Patterson. 1967. Six cases of acrylamide poisoning. Br. Med. J. 4: 34-138. Garvin, P. J., M. E. Lewandowski, and R. F. Wallin. 1976. A teratologic evaluation of plasma-soluble extracts of polyvinyl chloride plastics. (Abstract 644.) Pharmacologist 18:231. Gibel, W., K. Lohs, G.-P. Wildner, and D. Ziebarth. 1971. Tierexperimentelle Unter- schungen uber die hepatotoxische und kanzerogene Wirkung phosphororganischer Ver- bindungen. Arch. Geschwulstforsch. 37:303-312. Gibel, W., K. Lohs, G.-P. Wildner, D. Ziebarth, and R. Stieglitz. 1973. Uber die kan- zerogene, hamatotoxische und hepatotoxische Wirkung pestizider organischer Phospho- roverbindungen. Arch. Geschwulstforsch. 41:311-328. Gilot-Delhalle, J., A. Colizzi, J. Moutschen, and M. Moutschen-Dahmen. 1983. Muta- genicity of some organophosphorus compounds at the ade6 locus of Schizosaccharomyces pombe. Mutat. Res. 117:139-148. Glass, R. I., R. N. Lyness, D. C. Mengle, K. E. Powell, and E. Kahn. 1979. Sperm count depression in pesticide applicators exposed to dibromochloropropane. Am. J. Epidemiol. 109:346-351. Goldstein, J. A., M. Friesen, R. E. Linder, P. Hickman, J. R. Hass, and H. Bergman. 1977. Effects of pentachlorophenol on hepatic drug-metabolizing enzymes and porphyria related to contamination with chlorinated dibenzo-p-dioxins and dibenzofurans. Biochem. Pharmacol. 26:1549-1557. Gordon, D. 1956. How dangerous is pentachlorophenol? Med. J. Aust. 43:485-488. Gosselin, R. E., R. P. Smith, and H. C. Hodge. 1984. Sulfallate. P. II-313 in Clinical Toxicology of Commercial Products, 5th ed. Williams & Wilkins, Baltimore. Gotz, R., L. R. Schwarz, and H. Greim. 1980. Effects of pentachlorophenol and 2,4,6- trichlorophenol on the disposition of sulfobromophthalein and respiration of isolated liver cells. Arch. Toxicol. 44:147-155. Gray, L. E., Jr., and R. J. Kavlock. 1983. The effects of the herbicide 2,4-dichlorophenyl- p-nitrophenyl ether (NIT) on serum thyroid hormones in adult female mice. Toxicol. Lett. 15:231-235. Gray, L. E., Jr., R. J. Kavlock, N. Chernoff, J. Ferrell, J. McLamb, and J. Ostby. 1982. Prenatal exposure to the herbicide 2,4-dichlorophenyl-p-nitrophenyl ether destroys the rodent Harderian gland. Science 215:293-294. Gray, T. J. B., and J. A. Beamand. 1984. Effect of some phthalate esters and other testicular toxins on primary cultures of testicular cells. Food Chem. Toxicol. 22:123- 131. Gray, T. J. B., and K. R. Butterworth. 1980. Testicular atrophy produced by phthalate esters. Arch. Toxicol. (Suppl. 4):452-455. Gray, T. J. B., and S. D. Gangolli. 1986. Aspects of the testicular toxicity of phthalate esters. Environ. Health Perspect. 65:229-235. Gray, T. J. B., K. R. Butterworth, I. F. Gaunt, P. Grasso, and S. D. Gangolli. 1977. Short-term toxicity study of di-(2-ethylhexyl) phthalate in rats. Food Cosmet. Toxicol. 15:389-399. Gray, T. J. B., J. A. Beamand, B. G. Lake, J. R. Foster, and S. D. Gangolli. 1982a. Peroxisome proliferation in cultured rat hepatocytes produced by clofibrate and phthalate ester metabolites. Toxicol. Lett. 10:273-279.

420 DRINKING WATER AND HEALTH Gray, T. J. B., I. R. Rowland, P. M. D. Foster, and S. D. Gangolli. 1982b. Species differences in the testicular toxicity of phthalate esters. Toxicol. Lett. 11:141-147. Greichus, Y. A., G. W. Libal, and D. D. Johnson. 1979. Diagnosis and physiologic effects of pentachlorophenols on young pigs. Part I. Effects of purified pentachlorophenol. Bull. Environ. Contam. Toxicol. 23:418-422. Grimm, H.-G., B. Schellmann, K.-H. Schaller, and K. Gossler. 1981. Pentachlorophenol concentrations in tissues and bodyfluids of normal persons. Zentralbl. Bakteriol., Mikrobiol. Hyg., Abt. 1, Orig. B 174:77-90. (In German; English abstract) Grzywa, Z., and E. Rudzki. 1981. Dermatitis from dichloropropane. Contact Derm. 7:151- 153. Hanna, S., K. Basmy, O. Selim, S. M. Shoeb, and A. Y. Awny. 1966. Effects of administration of an organo-phosphorus compound as an antibilharzial agent, with special reference to plasma cholinesterase. Br. Med. J. 1: 1390- 1392. Hardin, B. D., G. P. Bond, M. R. Sikov, F. D. Andrew, R. P. Beliles, and R. W. Niemeier. 1981. Testing of selected workplace chemicals for teratogenic potential. Scand. J. Work Environ. Health 7(Suppl. 4):66-75. Harris, R. S., H. C. Hodge, E. A. Maynard, and H. J. Blanchet, Jr. 1956. Chronic oral toxicity of 2-ethylhexyl phthalate in rats and dogs. AMA Arch. Ind. Health 13:259-264. Hashimoto, K., and W. N. Aldridge. 1970. Biochemical studies on acrylamide, a neurotoxic agent. Biochem. Pharmacol. 19:2591-2604. Hassan, A., and S. M. A. D. Zayed. 1965. Metabolism of organophosphorus insecticides. III. Fate of the methyl groups of dipterex in vivo. Can. J. Biochem. 43:1271-1275. Hassan, A., S. M. A. D. Zayed, and S. Hashish. 1965. Metabolism of organophosphorus insecticides. VI. Mechanism of detoxification of Dipterex in the rat. Biochem. Phar- macol. 14:1692-1694. Hawton, D., and E. H. Stobbe. 1971. The fate of nitrofen in rape, redroot pigweed, and green foxtail. Weed Sci. 19:555-558. Heppel, L. A., P. A. Neal, B. Highman, and V. T. Porterfield. 1946. Toxicology of 1,2- dichloropropane (propylene dichloride). I. Studies on effects of daily inhalations. J. Ind. Hyg. Toxicol. 28:1-8. Heppel, L. A., B. Highman, and E. G. Peake. 1948. Toxicology of 1,2-dichloropropane (propylene dichloride). IV. Effects of repeated exposures to a low concentration of the vapor. J. Ind. Hyg. Toxicol. 30:189-191. Hesterberg, T. W., and J. C. Barrett. 1984. Dependence of asbestos- and mineral dust- induced transformation of mammalian cells in culture on fiber dimension. Cancer Res. 44:2170-2180. Hiatt, C. W., W. T. Haskins, and L. Olivier. 1960. The action of sunlight on sodium pentachlorophenate. Am. J. Trop. Med. Hyg. 9:527-531. Hierons, R., and M. K. Johnson. 1978. Clinical and toxicological investigations of a case of delayed neuropathy in man after acute poisoning by an organophosphorus pesticide. Arch. Toxicol. 40:279-284. Highman, B., and L. A. Heppel. 1946. Toxicology of 1,2-dichloropropane (propylene dichloride). III. Pathologic changes produced by a short series of daily exposures. Arch. Pathol. 42:525-534. Hine, C. H., H. H. Anderson, H. D. Moon, J. K. Kodama, M. Morse, and N. W. Jacobsen. 1953. Toxicology and safe handling of CBP-55 (technical 1-chloro-3-bromopropene-1). AMA Arch. Ind. Hyg. Occup. Med. 7:118-136. Hinkle, D. K. 1973. Fetotoxic effects of pentachlorophenol in the golden Syrian hamster. (Abstract 42.) Toxicol. Appl. Pharmacol. 25:455.

Toxicity of Selected Contaminants 42 ~ Hoak, R. D. 1957. The causes of tastes and odors in drinking water. Water Sewage Works 104:243-247. Hodge, H. C. 1943. Acute toxicity for rats and mice of 2-ethyl hexanol and 2-ethyl hexyl phthalate. Proc. Soc. Exp. Biol. Med. 53:20-23. Hodgson, J. R., B. C. Myhr, M. McKeon, and D. J. Brusick. 1982. Evaluation of di(2- ethylhexyl)phthalate and its major metabolites in the orimarv rat hepatocvte unscheduled DNA synthesis assay. (Abstract Dg-3.) Environ. Mutagen. 4:388. Holmstedt, B., I. Nordgren, M. Sandoz, and A. Sundwall. 1978. Metrifonate: Summary of toxicological and pharmacological information available. Arch. Toxicol 41:3-29. Hutson, D. H., J. A. Moss, and B. A. Pickering. 1971.The excretion and retention of components of the soil fumigant D-D and their metabolites in the rat. Food Cosmet. Toxicol. 9:677-680. IARC (International Agency for Research on Cancer). 1976. Diallate. Pp. 69-75 in IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Man. Vol. 12. Some Carbamates, Thiocarbamates, and Carbazides. International Agency for Research on Cancer, Lyon, France. IARC (International Agency for Research on Cancer). 1979. Pentachlorophenol. Pp. 303- 325 in IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Vol. 20. Some Halogenated Hydrocarbons. International Agency for Research on Cancer, Lyon, France. IARC (International Agency for Research on Cancer). 1982. Di(2-ethylhexyl)phthalate. Pp. 269-294 in IARC Monographs on the Evaluation of the Carcinogenic Risk of Chem- icals to Humans. Vol. 29. Some Industrial Chemicals and Dyestuffs. International Agency for Research on Cancer, Lyon, France. IARC (International A~encY for Research on Cancer). 1983. IARC Mono~ranhs on the r ~ ~ -~ --r-~~~ -~ ~ Evaluation of the Carcinogenic Risk of Chemicals to Humans. Vol. 30. Miscellaneous Pesticides. International Agency for Research on Cancer, Lyon, France. Igisu, H., I. Goto, Y. Kawamura, M. Kato, K. Izumi, and Y. Kuroiwa. 1975. Acrylamide encephaloneuropathy due to well water pollution. J. Neurol. Neurosurg. Psychiatry 38:581-584. Ikeda, G. J., P. P. Sapienza, J. L. Couvillion, T. M. Farber, and E. J. van Loon. 1980. Comparative distribution, excretion and metabolism of di-(2-ethylhexyl) phthalate in rats, dogs and miniature pigs. Food Cosmet. Toxicol. 18:637-642. Innes, J. R. M., B. M. Ulland, M. G. Valerio, L. Petrucelli, L. Fishbein, E. R. Hart, A. J. Pallotta, R. R. Bates, H. L. FaLk, J. J. Gart, M. Klein, I. Mitchell, and J. Peters. 1969. Bioassay of pesticides and industrial chemicals for tumorigenicity in mice: A preliminary note. J. Natl. Cancer Inst. 42:1101-1114. Ivanova, T. P. 1967. The maximum permissible concentration of nitraphen in the air of the working zone. Hyg. Sanit. 32:20-26. Jaeger, R. J., and R. J. Rubin. 1972. Migration of a phthalate ester plasticizer from polyvinyl chloride blood bags into stored human blood and its localization in human tissues. N. Engl. J. Med. 287:1114-1118. Jewsbury, J. M., M. J. Cooke, and M. C. Weber. 1977. Field trial of metrifonate in the treatment and prevention of schistosomiasis infection in man. Ann. Trop. Med. Parasitol. 71 :67-83. Johnson, K. A., S. J. Gorzinski, K. M. Bodner, and R. A. Campbell. 1984. Acrylamide: A Two-Year Drinking Water Chronic Toxicity-Oncogenicity Study in Fischer 344 Rats. Dow Chemical USA, Midland, Mich. 258 pp. Johnson, M. K. 1970. Organophosphorus and other inhibitors of brain 'neurotoxic esterase' and the development of delayed neurotoxicity in hens. Biochem. J. 120:523-531.

422 DRINKING WATER AND HEALTH Johnson, M. K. 1981. Delayed neurotoxicity. Do trichlorphon and/or dichlorvos cause delayed neuropathy in man or in test animals? Acta Pharmacol. Toxicol. 49(Suppl. V):87- 98. Johnson, R. L., P. J. Gehring, R. J. Kociba, and B. A. Schwetz. 1973. Chlorinated dibenzodioxins and pentachlorophenol. Environ. Health Perspect. 5:171-175. Jones, A. R., and J. Gibson. 1980. 1,2-Dichloropropane: Metabolism and fate in the rat. Xenobiotica 10:835-846. Kale, P. G., and J. W. Baum. 1983. Sensitivity of Drosophila melanogaster to low concentrations of gaseous mutagens. IV. Mutations in embryonic spermatogonia. Mutat. Res. 113: 135-143. Kalow, W. 1965. Genetic factors in relation to drugs. Annul Rev. Pharmacol. 5:9-26. Kang, Y. J., L. Zolna, and J. M. Manson. 1985. Strain differences in response of rats to the teratogen nitrofen (2,4-dichloro-4'-nitrodiphenyl ether). (Abstract P49.) Teratology 31 :54A-SSA. Kapp, R. W., Jr., D. J. Picciano, and C. B. Jacobson. 1979. Y-chromosomal nondisjunction in dibromochloropropane-exposed workmen. Mutat. Res. 64:47-51. Kato, Y., K. Sato, O. Matano, and S. Goto. 1980a. Alkylation of cellular macromolecules by reactive metabolic intermediate of DBCP. J. Pestic. Sci. 5:45-53. Kato, Y., K. Sato, T. Harada, S. Maki, O. Matana, and S. Goto. 1980b. Correlation between macromolecular binding of DBCP-metabolite and pathogenicity of necrosis. J. Pestic. Sci. 5:81-88. Kavlock, R. J., and J. A. Gray. 1983. Morphometric, biochemical, and physiological assessment of perinatally induced renal dysfunction. J. Toxicol. Environ. Health 11:1- 13. Kavlock, R. J., and L. E. Gray, Jr. 1983. Postnatal evaluation of morphological and functional effects of prenatal exposure to nitrofen in the Long-Evans rat. J. Toxicol. Environ. Health 11:679-690. Kavlock, R. J., N. Chernoff, E. Rogers, D. Whitehouse, B. Carver, J. Gray, and K. Robinson. 1982. An analysis of fetotoxicity using biochemical endpoints of organ differentiation. Teratology 26:183-194. Kerkvliet, N. I., L. Baecher-Steppan, and J. A. Schmitz. 1982a. Immunotoxicity of pen- tachlorophenol (PCP): Increased susceptibility to tumor growth in adult mice fed technical PCP-contaminated diets. Toxicol. Appl. Pharmacol. 62:55-64. Kerkvliet, N. I., L. Baecher-Steppan, A. T. Claycomb, A. M. Craig, and G. G. Sheggeby. 1982b. Immunotoxicity of technical pentachlorophenol (PCP-T): Depressed humoral immune responses to T-dependent and T-independent antigen stimulation in PCP-T exposed mice. Fund. Appl. Toxicol. 2:90-99. Kharrazi, M., G. Potashnik, and J. R. Goldsmith. 1980. Reproductive effects of dibromo- chloropropane. Isr. J. Med. Sci. 16:403-406. Kimbrough, R. D., and R. E. Linder. 1978. The effect of technical and purified pentachlo- rophenol on the rat liver. Toxicol. Appl. Pharmacol. 46:151-162. Kimbrough, R. D., T. B. Gaines, and R. E. Linder. 1974. 2,4-Dichloro-phenyl-p- nitrophenyl ether (TOK). Effects on the lung maturation of rat fetus. Arch. Environ. Health 28:316-320. Kirby, P. E., R. F. Pizzarello, T. E. Lawlor, S. R. Haworth, and J. R. Hodgson. 1983. Evaluation of di-(2-ethylhexyl)phthalate and its major metabolites in the Ames test and L5178Y mouse lymphoma mutagenicity assay. Environ. Mutagen. 5:657-663.

Toxicity of Selected Contaminants 423 Klemmer, H. W., L. Wong, M. M. Sato, E. L. Reichert, R. J. Korsak, and M. N' Rashad. 1980. Clinical findings in workers exposed to pentachlorophenol. Arch. Environ. Con- tam. Toxicol. 9:715-725. Kluwe, W. M. 1981a. Acute toxicity of 1,2-dibromo-3-chloropropane in the F344 male rat. I. Dose-response relationships and differences in routes of exposure. Toxicol. Appl. Pharmacol. 59:71-83. Kluwe, W. M. 1981b. Acute toxicity of 1,2-dibromo-3-chloropropane in the F344 male rat. II. Development and repair of the renal, epididymal, testicular, and hepatic lesions. Toxicol. Appl. Pharmacol. 59:84-95. Kluwe, W. M. 1983. Chemical modulation of 1,2-dibromo-3-chloropropane toxicity. Tox- icology 27:287-299. Kluwe, W. M., R. McNish, K. Smithson, and J. B. Hook. 1981. Depletion by 1,2- dibromoethane, 1,2-dibromo-3-chloropropane, tris(2,3-dibromopropyl)phosphate, and hexachloro-1,3-butadiene of reduced non-protein sulfhydryl groups in target and non- target organs. Biochem. Pharmacol. 30:2265-2271. Kluwe, W. M., A. Greenwell, and F. Harrington. 1982. Relationship of tissue nonprotein sulfhydryls to the acute toxic effects of 1,2-dibromo-3-chloropropane. J. Pharmacol. Exp. Ther. 220:399-405. Kluwe, W. M., B. N. Gupta, and J. C. Lamb IV. 1983. The comparative effects of 1,2- dibromo-3-chloropropane (DBCP) and its metabolites, 3-chloro-1,2-propaneoxide (epichlorohydrin), 3-chloro-1,2-propanediol (alphachlorohydrin), and oxalic acid, on the urogenital system of male rats. Toxicol. Appl. Pharmacol. 70:67-86. Knudsen, I., H. G. Verschuuren, E. M. Den Tonkelaar, R. Kroes, and P. F. W. Helleman. 1974. Short-term toxicity of pentachlorophenol in rats. Toxicology 2:141-152. Kornbrust, D. J., T. R. Barfknecht, P. Ingram, and J. D. Shelburne. 1984. Effect of di(2- ethylhexyl)phthalate on DNA repair and lipid peroxidation in rat hepatocytes and on metabolic cooperation in Chinese hamster V-79 cells. J. Toxicol. Environ. Health 13:99- 116. Kowalski, B., E. B. Brittebo, and I. Brandt. 1985. Epithelial binding of 1,2-dibromoethane in the respiratory and upper alimentary tracts of mice and rats. Cancer Res. 45:2616- 2625. Kozumbo, W. J., R. Kroll, and R. J. Rubin. 1982. Assessment of the mutagenicity of phthalate esters. Environ. Health Perspect. 45:103-109. Kuperman, A. S. 1958. Effects of acrylamide on the central nervous system of the cat. J. Pharmacol. Exp. Ther. 123:180-192. Kutz, F. W., R. S. Murphy, and S. C. Strassman. 1978. Survey of pesticide residues and their metabolites in urine from the general population. Pp. 363-369 in K. R. Rao, ed. Pentachlorophenol: Chemistry, Pharmacology, and Environmental Toxicology. Plenum, New York. Lake, B. G., P. G. Brantom, S. D. Gangolli, K. P. Butterworth, and P. Grasso. 1976. Studies on the effects of orally administered di-(2-ethylhexyl) phthalate in the ferret. Toxicology 6:341-356. Lande, S. S., S. J. Bosch, and P. H. Howard. 1979. Degradation and leaching of acrylamide in soil. J. Environ. Qual. 8:133-137. Lantz, G. D., G. R. Cunningham, C. Huckins, and L. I. Lipshultz. 1981. Recovery from severe oligospermia after exposure to dibromochloro-propane. Fertil. Steril. 35:46-53. Larsen, R. V., G. S. Born, W. V. Kessler, S. M. Shaw, and D. C. Van Sickle. 1975. Placental transfer and teratology of pentachlorophenol in rats. Environ. Lett. 10(2): 121- 128.

424 DRINKING WATER AND H"LTH Lebrun, A., and C. Cerf. 1960. Note preliminaire sur la toxicite pour l'homme d'un insecticide organophosphore (Dipterex). WHO Bull. 22:579-582. Lee, I. P., and K. Suzuki. 1979. Induction of unscheduled DNA synthesis in mouse germ cells following 1-2-dibromo-3-chloropropane (DBCP) exposure. Mutat. Res. 68:169- 173. Leighty, E. G., and A. F. Fentiman, Jr. 1982. Conjugation of pentachlorophenol to palmitic acid by liver microsomes. Bull. Environ. Contam. Toxicol. 28:329-333. Letz, G. A., S. M. Pond, J. D. Osterloh, R. L. Wade, and C. E. Becker. 1984. Two fatalities after acute occupational exposure to ethylene dibromide. J. Am. Med. Assoc. 252:2428-2431. Levine, R. J., M. J. Symons, S. A. Balogh, T. H. Milby, and M. D. Whorton. 1981. A method for monitoring the fertility of workers. 2. Validation of the method among workers exposed to dibromochloropropane. J. Occup. Med. 23:183-188. Levine, R. J., P. B. Blunden, D. DalCorso, T. B. Starr, and C. E. Ross. 1983. Superiority of reproductive histories to sperm counts in detecting infertility at a dibromochloropropane manufacturing plant. J. Occup. Med. 25:591-597. Lewandowski, M. E., J. Fernandes, and T. S. Chen. 1980. Assessment of the teratogenic potential of plasma-soluble extracts of diethylhexyl phthalate plasticized polyvinyl chlo- ride plastics in rats. Toxicol. Appl. Pharmacol. 54:141-147. Lindgren, A., N. G. Lindquist, A. Lyden, T. Olsson, and S. Ullberg. 1982. A whole body autoradiographic study on the distribution of ~4C-labelled di-(2-ethylhexyl)phthalate in mice. Toxicology 23:149-158. Lipshultz, L. I., C. E. Ross, D. Whorton, T. Milby, R. Smith, and R. E. Joyner. 1980. Dibromochloropropane and its effect on testicular function in man. J. Urol. 124:464- 468. Litwin, Y. J., N. N. Hantzsche, and N. A. George. 1983. Groundwater Contamination by Pesticides: A California Assessment. Report submitted to the California Water Re- sources Control Board by Ramlit Associates, Inc. 209 pp. Lutz, W. K. 1986. Investigation of the potential for binding of di(2-ethylhexyl) phthalate (DEHP) to rat liver DNA in vivo. Environ. Health Perspect. 65:267-269. Machemer, L. 1981. Chronic toxicity of metrifonate. Acta Pharmacol. Toxicol. 49(Suppl. V): 15-28. MacWilliams, D. C. 1978. Acrylamide. Pp. 298-311 in Kirk-Othmer Encyclopedia of Chemical Technology, 3rd ed. Vol. I. John Wiley, New York. Malcolm, A. R., L. J. Mills, and E. J. McKenna. 1983. Inhibition of metabolic cooperation between Chinese hamster V79 cells by tumor promoters and other chemicals. Ann. N.Y. Acad. Sci. 407:448-450. Manson, J. M., T. Brown, and D. M. Baldwin. 1984. Teratogenicity of nitrofen (2,4- dichloro-4'-nitrodiphenyl ether) and its effects on thyroid function in the rat. Toxicol. Appl. Pharmacol. 73:323-335. Marquez Mayaudon, E. 1978. 1,2 Dibromo-3-chlor-propane (DBCP) nematocide with sterilizing action on man. Salud Publica Mex. 20: 195-200. (In Spanish; English abstract) McCollister, D. D., F. Oyen, and V. K. Rowe. 1964. Toxicology of acrylamide. Toxicol. Appl. Pharmacol. 6:172-181. McConnell, E. E., J. A. Moore, B. N. Gupta, A. H. Rakes, M. I. Luster, J. A. Goldstein, J. K. Haseman, and C. E. Parker. 1980. The chronic toxicity of technical and analytical pentachlorophenol in cattle. I. Clinicopathology. Toxicol. Appl. Pharmacol. 52:468-490. McGavack, T. H., L. J. Boyd, F. V. Piccione, and R. Terranova. 1941. Acute and chronic intoxications with sodium pentachlorophenate in rabbits. J. Ind. Hyg. Toxicol. 23:239- 251.

Toxicity of Selected Contaminants 425 Meerman, J. H. N., and G. J. Mulder. 1981. Prevention of the hepatotoxic action of N- hydroxy-2-acetylaminofluorene in the rat by inhibition of N-O-sulfation by pentachlo- rophenol. Life Sci. 28:2361-2365. Mehrle, P. M., and F. L. Mayer. 1976. Di-2-ethylhexyl phthalate: Residue dynamics and biological effects in rainbow trout and fathead minnows. Trace Subst. Environ. Health 10:519-524. Menon, J. A. 1958. Tropical hazards associated with the use of pentachlorophenol. Br. Med.J.1:1156-1158. Mercier, M. 1976. Pentachlorophenol. Pp. 2-18 in Preparatory Study for Establishing Criteria (Exposure/Effect Relationships) for Humans on Organochlorine Compounds, i.e., Pesticides. 2nd Series. Doc. No. V/F/2500/76 e. Commission of the European Communities, Luxembourg. Metcalf, R. L., T. R. Fukuto, and R. B. March. 1959. Toxic action of Dipterex and DDVP to the house fly. J. Econ. Entomol. 52:44-49. Milby, T. H., and D. Whorton. 1980. Epidemiological assessment of occupationally related, chemically induced sperm count suppression. J. Occup. Med. 22:77-82. Millar, M. J., P. V. Elcoate, M. I. Fischer, and C. A. Mawson. 1960. Effect of testosterone and gonadotrophin injections on the sex organ development of zinc-deficient male rats. Can. J. Biochem. Biophysiol. 38:1457-1466. Miller, M. J., D. E. Carter, and I. G. Sipes. 1982. Pharmacokinetics of acrylamide in Fischer-344 rats. Toxicol. Appl. Pharmacol. 63:36-44. Miller, M. J., M. S. Miller, and C. A. McQueen. 1984. Lack of genotoxicity of acrylamide using the hepatocyte primary culture (HPC)/DNA repair test. (Abstract 138.) Toxicologist 4:35. Miller, M. S., M. J. Miller, T. F. Burks, and I. G. Sipes. 1983. Altered retrograde axonal transport of nerve growth factor after single and repeated doses of acrylamide in the rat. Toxicol. Appl. Pharmacol. 69:96-101. Milman, H. A., J. M. Ward, and K. C. Chu. 1978. Pancreatic carcinogenesis and naturally occurring pancreatic neoplasms of rats and mice in the NCI Carcinogenesis Testing Program. J. Environ. Pathol. Toxicol. 1:829-840. Mirsalis, J. C., and B. E. Butterworth. 1980. Detection of unscheduled DNA synthesis in hepatocytes isolated from rats treated with genotoxic agents: An in vivo-in vitro assay for potential carcinogens and mutagens. Carcinogenesis 1 :621-625. Mitchell, A. M., J. W. Bridges, and C. R. Elcombe. 1984. Factors influencing peroxisome proliferation in cultured rat hepatocytes. Arch. Toxicol. 55:239-246. Mitchell, F. E., S. C. Price, P. Grasso, R. H. Hinton, and J. W. Bridges. 1984. Effects of all-(2-ethyl hexyl)phthalate on male and ferrale rats. (Abstract.) Paper presented at the International Conference on Phthalic Acid Esters, August 6-7, 1984, University of Surrey, Guilford, Surrey, England. Miyazawa, S., S. Furuta, T. Osumi, and T. Hashimoto. 1980. Turnover of enzymes of peroxisomal p-oxidation in rat liver. Biochim. Biophys. Acta 630:367-374. Moody, D. E., M. S. Rao, and J. K. Reddy. 1977. Mitogenic effect in mouse liver induced by a hypolipidemic drug, nafenopin. Virchows Arch. B 23:291-296. Moody, D. E., J. L. James, G. A. Clawson, and E. A. Smuckler. 1981. Correlations among the changes in hepatic microsomal components after intoxication with aLkyl halides and other hepatotoxins. Mol. Pharmacol. 20:685-693. Moody, D. E., G. A. Clawson, C. H. Woo, and E. A. Smuckler. 1982a. Cellular distri- bution of cytochrome P-450 loss in rats of different ages treated with alkyl halides. Toxicol. Appl. Pharmacol. 66:278-289.

426 DRINKING WATER AND H"LTH Moody, D. E., G. A. Clawson, and E. A. Smuckler. 1982b. The integrity of liver protein synthesis in male rats treated with 1,2-dibromo-3-chloropropane. Toxicol. Lett. 12:101- 108. Moreland, D. E., W. J. Blackmon, H. G. Todd, and F. S. Farmer. 1970. Effects of diphenylether herbicides on reactions of mitochondria and chloroplasts. Weed Sci. 18:636- 642. Morgade, C., A. Barquet, and C. D. Pfaffenberger. 1980. Determination of polyhaloge- nated phenolic compounds in drinking water, human blood serum and adipose tissue. Bull. Environ. Contam. Toxicol. 24:257-264. Moriya, M., T. Ohta, K. Watanabe, T. Miyazawa, K. Kato, and Y. Shirasu. 1983. Further mutagenicity studies on pesticides in bacterial reversion assay systems. Mutat. Res. 116:185-216. Morpurgo, G., F. Aulicino, M. Biganami, L. Conti, and A. Velcich. 1977. Relationship between structure and mutagenicity of dichlorvos and other pesticides. Atti Accad. Naz. Lincei C1. Sci. Fis. Mat. Nat. Rend. 62:692-701. Morton, S. J. 1979. The Hepatic Effects of Dietary Di-2-ethylhexyl Phthalate. Ph.D. Thesis. The Johns Hopkins University. University Microfilms International, Ann Arbor, Mich- igan. 135 pp. [Abstract in Diss. Abstr. Int. B 40(09):4236.] Moutschen-Dahmen, J., M. Moutschen-Dahmen, and N. Degraeve. 1981. Metrifonate and dichlorvos: Cytogenetic investigations. Acta Pharmacol. Toxicol. 49(Suppl. V):29-39. Murphy, S. D. 1980. Pesticides. Pp. 357-408 in J. Doull, C. D. Klaassen, and M. O. Amdur, eds. Casarett and Doull's Toxicology. The Basic Science of Poisons, 2nd ed. Macmillan, New York. Nachtomi, E. 1980. Modulation of the mitotic action of ethylene dibromide. Chem.-Biol. Interact. 32:311-319. Nachtomi, E., and E. Alumot. 1972. Comparison of ethylene dibromide and carbon tetra- chloride toxicity in rats and chicks: Blood and liver levels; lipid peroxidation. Exp. Mol. Pathol. 16:71-78. Nachtomi, E., and D. S. R. Sarma. 1977. Repair of rat liver DNA in vivo damaged by ethylene dibromide. Biochem. Pharmacol. 26:1941-1945. Nakamura, Y., Y. Yagi, I. Tomita, and K. Tsuchikawa. 1979. Teratogenicity of di-(2- ethylhexyl)phthalate in mice. Toxicol. Lett. 4:113-117. NCI (National Cancer Institute). 1978a. Bioassay of Dibromochloropropane for Possible Carcinogenicity. Technical Report Series No. 28. DHEW Publication No. (NIH) 78- 828. U.S. Department of Health, Education, and Welfare, Washington, D.C. [98 pp.] NCI (National Cancer Institute). 1978b. Bioassay of 1,2-Dibromoethane for Possible Car- cinogenicity. Technical Report Series No. 86. DHEW Publication No. (NIH) 78-1336. U.S. Department of Health, Education, and Welfare, Washington, D.C. [127 pp.] NCI (National Cancer Institute). 1978c. Bioassay of Nitrofen for Possible Carcinogenicity. Technical Report Series No. 26. DHEW Publication No. (NIH) 78-826. U.S. Department of Health, Education, and Welfare, Washington, D.C. [101 pp.] NCI (National Cancer Institute). 1978d. Bioassay of Sulfallate for Possible Carcinogenicity. Technical Report Series No. 115. DHEW Publication No. (NIH) 78-1730. U.S. De- partment of Health, Education, and Welfare, Washington, D.C. [75 pp.] NCI (National Cancer Institute). 1979. Bioassay of Nitrofen for Possible Carcinogenicity. Technical Report Series No. 184. DHEW Publication No. (NIH) 79-1740. U.S. De- partment of Health, Education, and Welfare, Washington, D.C. [57 pp.] NCI (National Cancer Institute). 1980. Bioassay of a Mixture of 1,2,3,6,7,8- and 1,2,3,7,8,9- Hexachlorodibenzo-p-dioxins for Possible Carcinogenicity (Gavage Study). Technical

Toxicity of Selected Contaminants 427 Report Series No. 198. DHHS Publication No. (NIH) 80-1754. U.S. Department of Health and Human Services, Washington, D.C. 173 pp. Nikonorow, N., H. Mazur, and H. Piekacz. 1973. Effect of orally administered plasticizers and polyvinyl chloride stabilizers in the rat. Toxicol. Appl. Pharmacol. 26:253-259. NIOSH (National Institute for Occupational Safety and Health). 1977a. Criteria for a Recommended Standard .... Occupational Exposure to Ethylene Dibromide. DHEW (NIOSH) Publication No. 77-221. U.S. Department of Health, Education, and Welfare, Washington, D.C. 208 pp. NIOSH (National Institute for Occupational Safety and Health). 1977b. National Occu- pational Hazard Survey. Vol. III. Survey Analysis and Supplemental Tables. DHEW (NIOSH) Publication No. 78- 114. National Institute for Occupational Safety and Health, Cincinnati. 799 pp. (Available from the National Technical Information Service, Spring- f~eld, Va., as Publication No. PB-82-229881.) NIOSH (National Institute for Occupational Safety and Health). 1978a. A Recommended Standard for Occupational Exposure to Dibromochloropropane. DHEW (NIOSH) Pub- lication No. 78-115. U.S. Department of Health, Education, and Welfare, National Institute for Occupational Safety and Health, Cincinnati. 14 pp. NIOSH (National Institute for Occupational Safety and Health). 1978b. NIOSH Current Intelligence Bulletin 23. Ethylene Dibromide and Disulfiram Toxic Interaction. DHEW (NIOSH) Publication No. 78-145. U.S. Department of Health, Education, and Welfare, Washington, D.C. 6 pp. NIOSH (National Institute for Occupational Safety and Health). 1981. NIOSH Current Intelligence Bulletin 37. Ethylene Dibromide (EDB) (Revised). DHHS (NIOSH) Pub- lication No. 82-105. U.S. Department of Health, Education, and Welfare, Washington, D.C. 13 pp. Nitschke, K. D., R. J. Kociba, D. G. Keyes, and M. J. McKenna. 1981. A thirteen week repeated inhalation study of ethylene bromide in rats. Fund. Appl. Toxicol. 1:437-442. Nordgren, I., E. Bengtsson, B. Holmstedt, and B.-M. Pettersson. 1981. Levels of metrif- onate and dichlorvos in plasma and erythrocytes during treatment of schistosomiasis with Bilarcil@. Acta Pharmacol. Toxicol. 49(Suppl. V):79-86. NRC (National Research Council). 1977. Drinking Water and Health. National Academy of Sciences, Washington, D.C. 938 pp. NRC (National Research Council). 1980. Drinking Water and Health. Vol. 3. National Academy Press, Washington, D.C. 415 pp. NRC (National Research Council). 1982a. Anticholinesterases. Pp. 5-50 in Possible Long- Term Health Effects of Short-Term Exposure to Chemical Agents. Vol. 1. Anticholin- esterases and Anticholinergics. National Academy Press, Washington, D.C. NRC (National Research Council). 1982b. Drinking Water and Health. Vol. 4. National Academy Press, Washington, D.C. 299 pp. NRC (National Research Council). 1983. Drinking Water and Health. Vol. 5. National Academy Press, Washington, D.C. 157 pp. NTP (National Toxicology Program). 1982a. Carcinogenesis Bioassay of 1,2-Dibromo-3- chloropropane (CAS No. 96-12-8) in F344 Rats and B6C3F~ Mice (Inhalation Study). NTP Technical Report Series No. 206. NIH Publication No. 82-1762. U.S. Department of Health and Human Services, Washington, D.C. [190 pp.] NTP (National Toxicology Program). 1982b. Carcinogenesis Bioassay of 1,2-Dibromo- ethane (CAS No. 106-93-4) in F344 Rats and B6C3F~ Mice (Inhalation Study). NTP Technical Report Series No. 210. NIH Publication No. 82-1766. U.S. Department of Health and Human Services, Washington, D.C. [175 pp.]

428 DRINKING WATER AND HEALTH NTP (National Toxicology Program). 1982c. Carcinogenesis Bioassay of Di(2-ethyl- hexyl)adipate (CAS No. 103-23-1) in F344 Rats and B6C3F~ Mice (Feed Study). NTP Technical Report Series No. 212. NIH Publication No. 81-1768. U.S. Department of Health and Human Services, Washington, D.C. [131 pp.] NTP (National Toxicology Program). 1983a. Carcinogenicity Bioassay of 1,2-Dichloro- propane (Propylene Dichloride) (CAS No. 78-87-5) in F344/N Rats and B6C3F~ Mice (Gavage Studies). NTP Technical Report Series No. 263. NIH Publication No. 84-2519 (preliminary). U.S. Department of Health and Human Services, Washington, D.C. 180 PP NTP (National Toxicology Program). 1983b. Carcinogenesis Bioassay of Di(2-ethyl- hexyl)phthalate (CAS No. 117-81-7) in F344 Rats and B6C3F~ Mice (Feed Study). NTP Technical Report Series No. 217. NIH Publication No. 82-1773. U.S. Department of Health and Human Services, Washington, D.C. [141 pp.] NTP (National Toxicology Program). 1984a. Report of the NTP Ad Hoc Panel on Chemical Carcinogenesis Testing and Evaluation, Board of Scientific Counselors, National Tox- icology Program. U.S. Department of Health and Human Services, Washington, D.C. 280 pp. NTP (National Toxicology Program). 1984b. Review of Current DHHS, DOE, and EPA Research Related to Toxicology. NTP Publication No. 84-024. U.S. Department of Health and Human Services, Washington, D.C. 387 pp. NTP (National Toxicology Program). 1985a. Toxicology and Carcinogenesis Studies of Telone IN (Technical-Grade 1,3-Dichloropropene [CAS No. 542-75-6] Containing 1.0% Epichlorohydrin as a Stabilizer) in F344/N Rats and B6C3F~ Mice (Gavage Studies). NTP Technical Report Series No. 269. NIH Publication No. 85-2525. U.S. Department of Health and Human Services, Washington, D.C. 159 pp. NTP (National Toxicology Program). 1985b. Toxicology Research and Testing Program. Management Status Report. Produced from NTP Chemtrack System. Research Triangle Park, N.C. O'Hara, G. P., P. K. Chan, J. C. Harris, S. S. Burke, J. M. Smith, and A. W. Hayes. 1983. The effect of nitrofen [4-(2,4-dichlorophenoxy)nitrobenzene] on the reproductive performance of male rats. Toxicology 28:323-333. Ohta, T., N. Nakamura, M. Moriya, Y. Shirasu, and T. Kada. 1984. The SOS-function- inducing activity of chemical mutagens in Escherichia coli. Mutat. Res. 131:101-109. Oishi, S. 1984a. Effects of di-2-ethylhexyl phthalate on lipid composition of serum and testis in rats. Toxicol. Lett. 23:67-72. Oishi, S. 1984b. Testicular atrophy of rats induced by di-2-ethylhexyl phthalate: Effects of vitamin A and zinc concentrations in the testis, liver and serum. Toxicol. Lett. 20:75- 78. Oishi, S., and K. Hiraga. 1980a. Testicular atrophy induced by phthalic acid esters: Effect on testosterone and zinc concentrations. Toxicol. Appl. Pharmacol. 53:35-41. Oishi, S., and K. Hiraga. 1980b. Testicular atrophy induced by phthalic acid monoesters: Effects of zinc and testosterone concentrations. Toxicology 15:197-202. Oishi, S., and K. Hiraga. 1982. Distribution and elimination of di-2-ethylhexyl phthalate (DEHP) and mono-2-ethylhexyl phthalate (MEHP) after a single oral administration of DEHP in rats. Arch. Toxicol. 51: 149- 155. Oishi, S., and K. Hiraga. 1983. Testicular atrophy induced by di-2-ethylhexyl phthalate: Effect of zinc supplement. Toxicol. Appl. Pharmacol. 70:43-48. Olajos, E. J., I. Rosenblum, F. Coulston, and N. Strominger. 1979. The dose-response relationship of trichlorfon neurotoxicity in hens. Ecotoxicol. Environ. Safety 3:245-255.

Toxicity of Selected Contaminants 429 Olmstead, E. V. 1960. Pathological changes in ethylene dibromide poisoning. AMA Arch. Ind. Health 21:525-529. Osterloh, J., G. Letz, S. Pond, and C. Becker. 1983. An assessment of the potential testicular toxicity of 10 pesticides using the mouse-sperm morphology assay. Mutat. Res. 116:407-415. Osterloh, J. D., B. Cohen, W. Popendorf, and S. M. Pond. 1984. Urinary excretion of the N-acetyl cysteine conjugate of cis-1,3-dichloropropene by exposed individuals. Arch. Environ. Health 39:271-275. Osumi, T., and T. Hashimoto. 1978. Enhancement of fatty acyl-CoA oxidizing activity in rat liver peroxisomes by di-(2-ethylhexyl)-phthalate. J. Biochem. 83:1361-1365. Ott, M. G., H. C. Scharnweber, and R. R. Langner. 1980. Mortality experience of 161 employees exposed to ethylene dibromide in two production units. Br. J. Ind. Med. 37: 163-168. Packham, E. D., J. E. Thompson, C. I. Mayfield, W. E. Inniss, and J. Kruuv. 1981. Perturbation of lipid membranes by organic pollutants. Arch. Environ. Contam. Toxicol. 10:347-356. Palmer, J. S., M. Haufler, L. M. Hunt, J. C. Schlinke, and C. E. Gates. 1972. Chronic toxicosis of sheep from the organic herbicide di-allate. Am. J. Vet. Res. 33:543-546. Parker, C. M., W. B. Coate, and R. W. Voelker. 1982. Subchronic inhalation toxicity of 1,3-dichloropropene/1,2-dichloropropane (D-D) in mice and rats. J. Toxicol. Environ. Health 9:899-910. Pastoor, T., and R. J. Richardson. 1981. Blood dynamics of acrylamide in rats. (Abstract 191.) Toxicologist 1:53. Pastoor, T., W. Heydens, and R. J. Richardson. 1980. Time- and dose-related excretion of acrylamide metabolites in the urine of Fisher-344 rats. (Abstract P.176.) Toxicol. Lett. S1:215. Permanent Commission and International Association of Occupational Health, Subcom- mittee on Pesticides. 1972. Epidemiological toxicology of pesticide exposure: Report of an international workshop. Arch. Environ. Health 25:399-405. Pestova, A. G. 1966. Toxicity of diptal and avadex. Gig. Toksikol. Pestits. Klin. Otravlenii 4:166-169. (In Russian; English summary in Chem. Abstr. 70:2642e, 1969) Phillips, B. J., T. E. B. James, and S. D. Gangolli. 1982. Genotoxicity studies of di(2- ethylhexyl)phthalate and its metabolites in CHO cells. Mutat. Res. 102:297-304. Phillips, B. J., D. Anderson, and S. D. Gangolli. 1986. Studies on the genetic effects of phthalic acid esters on cells in culture. Environ. Health Perspect. 65:263-266. Pik, J., J. Braier, Z. Kalman, and M. Jozef. 1979. Determination of di(2-ethyl- hexyl)phthalate (DEHP) in blood preparations stored in poly(vinyl chloride) bags (Rum). Probl. Gematol. Pereliv. Krovi 24:49-51. (In Russian; English summary in Chem. Abstr. 91:27213n, 1979) Popp, J. A., L. K. Garvey, T. E. Hamm, Jr., and J. A. Swenberg. 1985. Lack of hepatic promotional activity by the peroxisomal proliferating hepatocarcinogen di(2-ethyl- hexyl)phthalate. Carcinogenesis 6:141-144. Potashnik, G., I. Yanai-Inbar, M. I. Sacks, and R. Israeli. 1979. Effect of dibromochlo- ropropane on human testicular function. Isr. J. Med. Sci. 15:438-442. Principe, P., E. Dogliotti, M. Bignami, R. Crebelli, E. Falcone, M. Fabrizi, G. Conti, and P. Comba. 1981. Mutagenicity of chemicals of industrial and agricultural relevance in Salmonella, Streptomyces and Aspergillus. J. Sci. Food Agric. 32:826-832. Prineas, J. 1969. The pathogenesis of dying-back polyneuropathies. Part II. An ultra structural study of experimental acrylamide intoxication in the cat. J. Neuropathol. Exp. Neurol. 28:598-621.

430 DRINKING WATER AND HEALTH Putman, D. L., W. A. Moore, L. M. Schechtman, and J. R. Hodgson. 1983. Cytogenetic evaluation of di-(2-ethylhexyl)phthalate and its major metabolites in Fischer 344 rats. Environ. Mutagen. 5:227-231. Rakhmatullaev, N. N. 1971. Hygienic characteristics of the nematocide Nemagon in relation to water pollution control. Hyg. Sanit. 36(1-3):344-348. Rao, K. S., J. D. Burek, F. J. Murray, J. A. John, B. A. Schwetz, J. E. Beyer, and C. M. Parker. 1982. Toxicologic and reproductive effects of inhaled 1,2-dibromo-3- chloropropane in male rabbits. Fund. Appl. Toxicol. 2:241-251. Raub, J. A., R. R. Mercer, R. J. Kavlock, and C. J. Setzer. 1983. Effects of prenatal nitrofen exposure on postnatal lung function in the rat. Pp. 119-134 in R. J. Kavlock and C. T. Grabowski, eds. Abnormal Functional Development of the Heart, Lungs, and Kidneys: Approaches to Functional Teratology. Alan R. Liss, New York. Reddy, J. K., and N. D. Lalwani. 1983. Carcinogenesis by hepatic peroxisome proliferators: Evaluation of the risk of hypolipidemic drugs and industrial plasticizers to humans. CRC Crit. Rev. Toxicol. 12:1-58. Reddy, J. K., M. S. Rao, and D. E. Moody. 1976. Hepatocellular carcinomas in acata- lasemic mice treated with Nafenopin, a hypolipidemic peroxisome proliferator. Cancer Res. 36:1211-1217. Reddy, J. K., D. L. Azarnoff, and C. R. Sirtori. 1978. Hepatic peroxisome proliferation: Induction by BR-931, a hypolipidemic analog of WY-14,643. Arch. Int. Pharmacodyn. 234:4- 14. Reddy, S., J. M. V. Oomen, and D. R. Bell. 1975. Metrifonate in urinary schistosomiasis: A field trial in northern Nigeria. Ann. Trop. Med. Parasitol. 69:73-76. Reel, J. R., A. D. Lawton, R. W. Tyl, and J. C. Lamb IV. 1985. Reproductive toxicity of dietary diethylhexyl phthalate (DEHP) in mouse breeding pairs. (Abstract 466.) Tox- icologist 5: 117. Reiner, E. 1981. Esterases in schistosomes: Reaction with substrates and inhititors. Acta Pharmacol. Toxicol. 49(Suppl. V):72-78. Renberg, L. 1974. Ion exchange technique for the determination of chlorinated phenols and phenoxy acids in organic tissue, soil, and water. Anal. Chem. 46:459-461. Reznik, G., H. Reznik-Schuller, J. M. Ward, and S. F. Stinson. 1980a. Morphology of nasal-cavity tumours in rats after chronic inhalation of 1,2-dibromo-3-chloropropane. Br. J. Cancer 42:772-781. Reznik, G., S. F. Stinson, and J. M. Ward. 1980b. Respiratory pathology in rats and mice after inhalation of 1,2-dibromo-3-chloropropane or 1,2-dibromoethane for 13 weeks. Arch. Toxicol. 46:233-240. Reznik, Y. B., and G. K. Sprinchan. 1975. Experimental data on gonadotoxic effect of Nemagon. Gig. Sanit. 975(6):101-102. (In Russian) Rhodes, C., T. Soames, M. D. Stonard, M. G. Simpson, A. J. Vernall, and C. R. Elcombe. 1984. The absence of testicular atrophy and in vivo and in vitro effects on hepatocyte morphology and peroxisomal enzyme activities in male rats following the administration of several alkanols. Toxicol. Lett. 21:103-109. Rhodes, C., T. C. Orton, I. S. Pratt, P. L. Batten, H. Bratt, S. J. Jackson, and C. R. Elcombe. 1986. Comparative pharmacokinetics and subacute toxicity of di(2-ethylhexyl) phthalate (DEPH) in rats and marmosets: Extrapolation of effects in rodents to man. Environ. Health Perspect. 65:299-308. Roberts, H. J. 1963. Aplastic anemia due to pentachlorophenol and tetrachlorophenol. South. Med. J. 56:632-635. Roberts, H. J. 1981. Aplastic anemia due to pentachlorophenol. N. Engl. J. Med. 305: 1650- 1651.

Toxicity of Selected Contaminants 43 ~ Robson, A. M., J. M. Kissane, N. H. Elvick, and L. Pundavela. 1969. Pentachlorophenol poisoning in a nursery for newborn infants. I. Clinical features and treatment. J. Pediatr. 75:309-316. Rosen, J. D., I. Schuphan, Y. Segall, and J. E. Casida. 1980. Mechanism for the mutagenic activation of the herbicide sulfallate. J. Agric. Food Chem. 28:880-881. Rosenkranz, H. S., and S. Rosenkranz. 1972. Reaction of DNA with phosphoric acid esters: Gasoline additive and insecticides. Experientia 28:386-387. Rowe, V. K., H. C. Spencer, D. D. McCollister, R. L. Hollingsworth, and E. M. Adams. 1952. Toxicity of ethylene dibromide determined on experimental animals. Arch. Ind. Hyg. Occup. Med. 6:158-173. Rowland, I. R., R. C. Cottrell, and J. C. Phillips. 1977. Hydrolysis of phthalate esters by the gastro-intestinal contents of the rat. Food Cosmet. Toxicol. 15: 17-21. Ruddick, J. A., and W. H. Newsome. 1979. A teratogenicity and tissue distribution study on dibromochloropropane in the rat. Bull. Environ. Contam. Toxicol. 21:483-487. Ruddick, J. A., D. C. Villeneuve, I. Chu, E. Nestmann, and D. Miles. 1981. An assessment of the teratogenicity in the rat and mutagenicity in Salmonella of mono-2-ethylhexyl phthalate. Bull. Environ. Contam. Toxicol. 27:181-186. Russfield, A. B. 1967. Pathology of the endocrine glands, ovary and testis of rats and mice. Pp. 391-467 in E. Cotchin and F. J. C. Roe, eds. Pathology of Laboratory Rats and Mice. F. A. Davis, Philadelphia. Saegusa, J., H. Hasegawa, and K. Kawai. 1982. Toxicity of 1,2-dibromo-3-chloropropane (DBCP). I. Histopathological examination of male rats exposed to DBCP vapour. Ind. Health 20:315-323. Sandifer, S. H., R. T. Wilkins, C. B. Loadholt, L. G. Lane, and J. C. Eldridge. 1979. Spermatogenesis in agricultural workers exposed to dibromochloropropane (DBCP). Bull. Environ. Contam. Toxicol. 23:703-710. Sangster, B., R. C. C. Wegman, and A. W. M. Hofstee. 1982. Non-occupational exposure to pentachlorophenol: Clinical findings and plasma-PCP-concentrations in three families. Human Toxicol. 1:123-133. Savolainen, H., and K. Pekari.- 1979. Neurochemical effects of peroral administration of technical pentachlorophenol. Res. Commun. Chem. Pathol. Pharmacol. 23:97-105. Schaumburg, H., J. Arezzo, and P. S. Spencer. 1982. Short-latency somatosensory evoked potentials in primates intoxicated with acrylamide: Implications for toxic neuropathies in man. (Abstract 490.) Toxicologist 2(1): 139. Schulz, C. O., and R. J. Rubin. 1973. Distribution, metabolism, and excretion of di-2- ethylhexyl phthalate in the rat. Environ. Health Perspect. 3:123-129. Schwetz, B. A., P. A. Keeler, and P. J. Gehring. 1974. The effect of purified and commercial grade pentachlorophenol on rat embyronal and fetal development. Toxicol. Appl. Pharmacol. 28:151-161. Schwetz, B. A., J. F. Quast, P. A. Keeler, C. G. Humiston, and R. J. Kociba. 1978. Results of two-year toxicity and reproduction studies on pentachlorophenol in rats. Pp. 301-309 in K. R. Rao, ed. Pentachlorophenol: Chemistry, Pharmacology, and Environ- mental Toxicology. Plenum, New York. Secchi, G. C., and L. Alessio. 1968. On the appearance of the mitochondrial isoenzyme of glutamic-oxalacetic transaminase in the serum of subject with acute liver disease from 1,2-dichloropropane and 1,2-dichloroethane poisoning. Med. Lav. 59:649-653. (In Ital- ian; English summary) Secchi, G. C., G. Chiappino, A. Lotto, and N. Zurlo. 1968. Actual chemical composition of the "commercial trichloroethylenes" and their liver toxicity. Clinical and enzymol- ogical studies. Med. Lav. 59:486-497. (In Italian; English summary)

432 DRINKING WATER AND HEALTH Seed, J. L. 1982. Mutagenic activity of phthalate esters in bacterial liquid suspension assays. Environ. Health Perspect. 45:111-114. Segerback, D., and L. Ehrenberg 1981. Alkylating properties of dichlorvos (DDVP). Acta Pharmacol. Toxicol. 49(Suppl. V):56-66. Senanayake, N., and M. K. Johnson. 1982. Acute polyneuropathy after poisoning by a new organophosphate insecticide. N. Engl. J. Med. 306:155-157. Seth, P. K., and M. Mushtaq. 1984. Age related sensitivity of rat testis towards di(2- ethylhexyl)phthalate. (Abstract.) Paper presented at the International Conference on Phthalic Acid Esters, August 6-7, 1984, University of Surrey, Guilford, Surrey, England. Shaffer, C. B., C. P. Carpenter, and H. F. Smyth, Jr. 1945. Acute and subacute toxicity of di(2-ethylhexyl) phthalate with note upon its metabolism. J. Ind. Hyg. Toxicol. 27: 130- 135. Shardina, R. A. 1972. Hygienic and toxicological assessment of nitrofen in water reservoirs. Med. Zh. Uzb. 5:26-28. (In Russian; English abstract 73-1449 in Health Aspects Pestic. 6:325) Sheldon, L. S., and R. A. Hites. 1979. Sources and movement of organic chemicals in the Delaware River. Environ. Sci. Technol. 13:574-579. Shiota, K., M. J. Chou, and H. Nishimura. 1980. Embryotoxic effects of di-2-ethylhexyl phthalate (DEHP) and di-n-butyl phthalate (DBP) in mice. Environ. Res. 22:245-253. Short, R. D., J. M. Winston, C.-B. Hong, J. L. Minor, C.-C. Lee, and J. Seifter. 1979. Effects of ethylene dibromide on reproduction in male and female rats. Toxicol. Appl. Pharmacol. 49:97-105. Sidorenko, G. I., V. R. Tsulaya, E. I. Korenevskaya, and T. I. Bonashevskaya. 1976. Methodological approaches to the study of the combined effect of atmospheric pollutants as illustrated by chlorinated hydrocarbons. Environ. Health Perspect. 13:111-116. Sikka, H. C., and P. Florczyk. 1978. Mutagenic activity of thiocarbamate herbicides in Salmonella typhimurium. J. Agric. Food Chem. 26:146-148. Singh, A. R., W. H. Lawrence, and J. Autian. 1972. Teratogenicity of phthalate esters in rats. J. Pharm. Sci. 61 :51 -55. Singh, A. R., W. H. Lawrence, and J. Autian. 1974. Mutagenic and antifertility sensitivities of mice to di-2-ethylhexyl phthalate (DEHP) and dimethoxyethyl phthalate (DMEP). Toxicol. Appl. Pharmacol. 29:35-46. Sjoberg, P., G. Montin, N.-G. Lindquist, and L. Ploen. 1984. Effects of repeated intra- venous infusions of the plasticizer di-(2-ethylhexyl) phthalate in male rats. (Abstract.) Paper presented at the International Conference on Phthalic Acid Esters, August 6-7, 1984, University of Surrey, Guilford, Surrey, England. Smith-Oliver, T., D. Loury, and B. E. Butterworth. 1985. Measurement of DNA repair and cell proliferation in hepatocytes from mice treated with di(2-ethylhexyl)phthalate (DEHP). (Abstract.) Environ. Mutagen. 7(Suppl. 3):71. Smyth, H. F., Jr., C. P. Carpenter, C. S. Weil, U. C. Pozzani, J. A. Striegel, and J. S. Nycum. 1969. Range-finding toxicity data: List VII. Am. Ind. Hyg. Assoc. J. 30:470- 476. Staples, R. E., and E. H. Goulding. 1979. Dipterex teratogenicity in the rat, hamster, and mouse when given by gavage. Environ. Health Perspect. 30:105-113. Steinberger, A., and E. Steinberger. 1977. The Sertoli cells. Pp. 371-399 in A. D. Johnson and W. R. Gomes, eds. The Testis. Vol. 4. Academic Press, New York. Stenchever, M. A., M. A. Allen, L. Jerominski, and R. V. Petersen. 1976. Effects of bis(2-ethyl hexyl) phthalate on chromosomes of human leukocytes and human fetal lung cells. J. Pharm. Sci. 65: 1648- 1651.

Toxicity of Selected Contaminants 433 Stolzenberg, S. J., and C. H. Hine. 1980. Mutagenicity of 2- and 3-carbon halogenated compounds in Salmonella/mammalian-microsome test. Environ. Mutagen. 2:59-66. Stone, L. C., and J. M. Manson. 1981. Effects of the herbicide 2,4-dichlorophenyl-p- nitrophenyl ether (nitrofen) on fetal lung development in rats. Toxicology 20:195-207. Storer, R. D., and R. B. Conolly. 1983. Comparative in vivo genotoxicity and acute hepatotoxicity of three 1,2-dihaloethanes. Carcinogenesis 4:1491-1494. Stott, W. T., R. H. Reitz, A. M. Schumann, and P. G. Watanabe. 1981. Genetic and nongenetic events in neoplasia. Food Cosmet. Toxicol. 19:567-576. Stott, W. T., P. E. Kastl, and M. J. McKenna. 1985. Inhalation pharmacokinetics of cis- and trans-1,3-dichloropropene in rats exposed to TELONE~ II vapors. (Abstract 440.) Toxicologist 5:110. Suzuki, K., and I. P. Lee. 1981. Induction of aryl hydrocarbon hydroxylase and epoxide hydrolase in rat liver, kidney, testis, prostate glands, and stomach by a potent nematocide, 1,2-dibromo-3-chloropropane. Toxicol. Appl. Pharmacol. 58:151-155. Takahashi, W., E. R. Reichert, G. C. Fung, and Y. Hokama. 1976. Acute phase proteins and pesticides exposure. Life Sci. 19:1645-1652. Talcott, R. E. 1981. Mutagenic impurities associated with 1,3-dichloropropene. (Abstract 149.) Toxicologist 1:41. Talcott, R. E., and J. King. 1984. Mutagenic impurities in 1,3-dichloropropene prepara- tions. J. Natl. Cancer Inst. 72:1113-1116. Taylor, P. 1980. Neuromuscular blocking agents. Pp. 220-234 in A. G. Gilman, L. S. Goodman, and A. Gilman, eds. The Pharmacological Basis of Therapeutics, 6th ed. Macmillan, New York. Teichmann, B., and F. Hauschild. 1978. Test of O,O-dimethyl(1-hydroxy-2,2,2-trichlo- roethyl)-phosphonate (Trichlorfon) for carcinogenic activity in mice by oral (oesophageal- gastric intubation), intraperitoneal and dermal application. Arch. Geschwulstforsch. 48:301- 307. (In German; English summary) Teichmann, B., and A. Schmidt. 1978. Test of O,O-dimethyl(1-hydroxy-2,2,2-trichloro- ethyl)phosphonate (Trichlorfon) for carcinogenic activity in Syrian golden hamsters (Mesocricetus auratus Waterhouse) by intraperitoneal administration. Arch. Geschwulst- forsch. 48:718-721. (In German; English summary) Teichmann, B., F. Hauschild, and A. Eckelmann. 1978. Test of O,O-dimethyl(1-hydroxy- 2,2,2-trichloroethyl)phosphonate (Trichlorfon) for carcinogenic activity in rats by oral (oesophageal-gastric intubation) and intraperitoneal administration. Arch. Geschwulst- forsch. 48:112-119. (In German; English summary) Ter Haar, G. 1980. An investigation of possible sterility and health effects from exposure to ethylene dibromide. Pp. 167-188 in B. Ames, P. Infante, and R. Reitz, eds. Ethylene Dichloride: A Potential Health Risk? Banbury Report 5. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Tezuka, H., N. Ando, R. Suzuki, M. Terahata, M. Moriya, and Y. Shirasu. 1980. Sister- chromatid exchanges and chromosomal aberrations in cultured Chinese hamster cells treated with pesticides positive in microbial reversion assays. Mutat. Res. 78:177-191. Thomas, J. A., end M. J. Thomas. 1984. Biologicaleffectsofdi-(2-ethylhexyl)phthalate and other phthalic acid esters. CRC Crit. Rev. Toxicol. 13:283-317. Thomas, J. A., T. D. Darby, R. F. Wallin, P. J. Garvin, and L. Martis. 1978. A review of the biological effects of di-(2-ethylhexyl) phthalate. Toxicol. Appl. Pharmacol. 45: 1- 27. Thomas, J. A., L. G. Schein, P. K. Gupta, R. E. McCafferty, P. R. Felice, and M. P. Donovan. 1979. Failure of monoethylhexyl phthalate to cause teratogenic effects in offspring of rabbits. Toxicol. Appl. Pharmacol. 51:523-528.

434 DRINKING WATER AND H"LTH Tofilon, P. J., R. P. Clement, and W. N. Piper. 1980. Inhibition of the biosynthesis of rat testicular heme by 1,2-dibromo-3-chloropropane. Biochem. Pharmacol. 29:2563- 2566. Tomita, I., Y. Nakamura, N. Aoki, and N. Inui. 1982. Mutagenic/carcinogenic potential of DEHP and MEHP. Environ. Health Perspect. 45:119-125. Torkelson, T. R., and F. Oyen. 1977. The toxicity of 1,3-dichloropropene as determined by repeated exposure of laboratory animals. Am. Ind. Hyg. Assoc. J. 38:217-223. Torkelson, T. R., S. E. Sadek, V. K. Rowe, J. K. Kodama, H. H. Anderson, G. S. Loquvam, and C. H. Hine. 1961. Toxicologic investigations of 1,2-dibromo-3-chloro- propane. Toxicol. Appl. Pharmacol. 3:545-559. Trevors, J. T., C. I. Mayfield, W. E. Inniss, and J. E. Thompson. 1981. Effect of phenolic antioxidants on the toxicity of pentachlorophenol in short-term bacterial bioassays. Bull. Environ. Contam. Toxicol. 27:433-439. Unger, T. M., D. Van Goethem, T. E. Shellenberger, J. Carter, J. Rollheiser, and G. Kuhn. 1983. Perinatal Toxicity of Nitrofen in Rats. MRI Project No. 4443-B(ll). Report to U.S. Environmental Protection Agency. Midwest Research Institute, Kansas City, Mo. 106 pp. U.S. International Trade Commission. 1981. Synthetic Organic Chemicals. United States Production and Sales, 1980. USITC Publication 1183. United States International Trade Commission, Washington, D.C. 327 pp. U. S. International Trade Commission. 1984. Imports of Benzenoid Chemicals and Products, 1983. USITC Publication 1548. United States International Trade Commission, Wash- ington, D.C. 101 pp. van Bao, T., I. Szabo, P. Ruzicska, and A. Czeizel. 1974. Chromosome aberrations in patients suffering acute organic phosphate insecticide intoxication. Humangenetik 24:33- 57. van Bladeren, P. J., D. D. Breimer, G. M. T. Rotteveel-Smijs, R. A. W. de Jong, W. Buijs, A. van der Gen, and G. R. Mohn. 1980. The role of glutathione conjugation in the mutagenicity of 1,2-dibromoethane. Biochem. Pharmacol. 29:2975-2982. van Bladeren, P. J., D. D. Breimer, G. M. T. Rotteveel-Smijs, P. de Knijff, G. R. Mohn, B. van Meeteren-Walchli, W. Buijs, and A. van der Gen. 1981. The relation between the structure of vicinal dihalogen compounds and their mutagenic activation via conju- gation to glutathione. Carcinogenesis 2:499-505. Vandekar, M. 1980. Minimizing occupational exposure to pesticides: Cholinesterase de- termination and organophosphorus poisoning. Residue Rev. 75:67-80. van Duuren, B. L., B. M. Goldschmidt, G. Loewengart, A. C. Smith, S. Melchionne, I. Seidman, and D. Roth. 1979. Carcinogenicity of halogenated olefinic and aliphatic hydrocarbons in mice. J. Natl. Cancer Inst. 63:1433-1439. Venable, J. R., C. D. McClimans, R. E. Flake, and D. B. Dimick. 1980. A fertility study of male employees engaged in the manufacture of glycerine. J. Occup. Med. 22:87-91. Vizethum, W., and G. Goerz. 1979. Induction of the hepatic microsomal and nuclear cytochrome P-450 system by hexachlorobenzene, pentachlorophenol and trichlorophenol. Chem.-Biol. Interact. 28:291-299. Vogel, E., and J. L. R. Chandler. 1974. Mutagenicity testing of cyclamate and some pesticides in Drosophila melanogaster. Experientia 30:621-623. Ward, J. M., R. A. Griesemer, and E. K. Weisburger. 1979. The mouse liver tumor as an endpoint in carcinogenesis tests. Toxicol. Appl. Pharmacol. 51:389-397. Ward, J. M., J. M. Rice, D. Creasia, P. Lynch, and C. Riggs. 1983. Dissimilar patterns of promotion by di(2-ethylhexyl)phthalate and phenobarbital of hepatocellular neoplasia initiated by diethylnitrosamine in B6C3F1 mice. Carcinogenesis 4:1021-1029.

Toxicity of Selected Contaminants 435 Ward, J. M., M. Ohshima, P. Lynch, and C. Riggs. 1984. Di(2-ethylhexyl) phthalate but not phenobarbital promotes N-nitrosodiethylamine-initiated hepatocellular proliferative lesions after short-term exposure in male B6C3F1 mice. Cancer Lett. 24:49-55. Ward, J. M., B. A. Diwan, M. Ohshima, H. Hu, H. M. Schuller, and J. M. Rice. 1986. Tumor-initiating and promoting activities of di(2-ethylhexyl) phthalate in viva and in vitro. Environ. Health Perspect. 65:279-291. Warren, J. R., N. D. Lalwani, and J. K. Reddy. 1982. Phthalate esters as peroxisome proliferator carcinogens. Environ. Health Perspect. 45:35-40. Waters, M. D., S. S. Sanded, V. F. Simmon, K. E. Mortelmans, A. D. Mitchell, T. A. Jorgenson, D. C. L. Jones, R. Valencia, and N. E. Garrett. 1982. Study of pesticide genotoxicity. Basic Life Sci. 21:275-326. Weatherholtz, W. M., K. S. Greenspun, and R. W. Kapp. 1979. Teratogenicity Studies in Rats, TOKEN E-25, TOKEN E-25 Solvent Control. Hazleton Laboratories America, Inc., Vienna, Va. Wegman, R. C. C., and A. W. M. Hofstee. 1979. Chlorophenols in surface waters of the Netherlands (1976-1977). Water Res. 13:651-657. Wegman, R. C. C., and H. H. van den Broek. 1983. Chlorophenols in river sediment in the Netherlands. Water Res. 17:227-230. Well, C. S., and C. P. Carpenter. 1968. TEMIK Sulfoxide. Results of Feeding in the Diet of Rats for Six Months and Dogs for Three Months. Special Report Number 31-141. Submitted to Union Carbide Corporation. Mellon Institute, Export, Pa. [51 pp.] Weinbach, E. C. 1954. The effect of pentachlorophenol on oxidative phosphorylation. J. Biol. Chem. 210:545-550. Weinbach, E. C., and J. Garbus. 1965. The interaction of uncoupling phenols with mi tochondria and with mitochondrial protein. J. Biol. Chem. 240:1811-1819. Weisburger, E. K., B. M. Ulland, J. Nam, J. J. Gart, and J. H. Weisburger. 1981. Carcinogenicity tests of certain environmental and industrial chemicals. J. Natl. Cancer Inst. 67:75-88. White, R. D., D. E. Carter, D. Earnest, and J. Mueller. 1980. Absorption and metabolism of three phthalate diesters by the rat small intestine. Food Cosmet. Toxicol. 18:383-386. White, R. D., I. G. Sipes, A. J. Gandolf~, and G. T. Bowden. 1981. Characterization of the hepatic DNA damage caused by 1,2-dibromoethane using the alkaline elusion tech- nique. Carcinogenesis 2:839-844. WHO (World Health Organization). 1972. Trichlorfon. Pp. 183-230 in 1971 Evaluations of Some Pesticide Residues in Food, the Monographs. WHO Pesticide Residues Series, No. 1. World Health Organization, Geneva. Whorton, D., R. M. Krauss, S. Marshall, and T. H. Milby. 1977. Infertility in male pesticide workers. Lancet 2:1259-1261. Whorton, D., T. H. Milby, R. M. Krauss, and H. A. Stubbs. 1979. Testicular function in DBCP exposed pesticide workers. J. Occup. Med. 21:161-166. Whorton, M. D., and D. E. Foliart. 1983. Mutagenicity, carcinogenicity and reproductive effects of dibromochloropropane (DBCP). Mutat. Res. 123:13-30. Whorton, M. D., and T. H. Milby. 1980. Recovery of testicular function among DBCP workers. J. Occup. Med. 22:177-179. Wilkinson, C. F., J. G. Babish, A. T. Lemley, and D. M. Soderlund. 1983. A Toxicological Evaluation of Aldicarb and Its Metabolites in Relation to the Potential Human Health Impact of Aldicarb Residues in Long Island Ground Water. A Report Prepared by a Committee from the Institute for Comparative and Environmental Toxicology. Cornell University, Ithaca, N.Y. [124 pp.]

436 DRINKING WATER AND H"LTH Williams, D. T., and B. J. Blanchf~eld. 1974. Retention, excretion and metabolism of di- (2-ethylhexyl) phthalate administered orally to the rat. Bull. Environ. Contam. Toxicol. 11:371-378. Williams, M. W., H. N. Fuyat, and O. G. Fitzhugh. 1959. The subacute toxicity of four organic phosphates to dogs. Toxicol. Appl. Pharmacol. 1:1-7. Williams, P. L. 1982. Pentachlorophenol, an assessment of the occupational hazard. Am. Ind. Hyg. Assoc. J. 43:799-810. Wills, J. H. 1972. The measurement and significance of changes in the cholinesterase activities of erythrocytes and plasma in man and animals. CRC Crit. Rev. Toxicol. 1: 153- 202. Windholz, M., S. Budavari, L. Y. Stroumtsos, and M. N. Fertig, eds. 1976. Propylene dichloride. P. 1017 in The Merck Index: An Encyclopedia of Chemicals and Drugs, 9th ed. Merck and Co., Rahway, N.J. Wiswesser, W. J., ed. 1976. P. 166 in Pesticide Index, 5th ed. The Entomological Society of America, College Park, Md. Wolkowski-Tyl, R., C. Jones-Price, M. C. Marr, and C. A. Kimmel. 1983a. Teratologic evaluation of diethylhexyl phthalate (DEHP) in CD-1 mice. (Abstract.) Teratology 27:84A- 85A. Wolkowski-Tyl, R., C. Jones-Price, M. C. Marr, and C. A. Kimmel. 1983b. Teratologic evaluation of diethylhexyl phthalate (DEHP) in Fischer 344 rats. (Abstract.) Teratology 27:85A. Wong, A. S., and D. G. Crosby. 1981. Photodecomposition of pentachlorophenol in water. J. Agric. Food Chem. 29:125-130. Wong, L. C. K., J. M. Winston, C. B. Hong, and H. Plotnick. 1982. Carcinogenicity and toxicity of 1,2-dibromoethane in the rat. Toxicol. Appl. Pharmacol. 63:155-165. Wong, O., H. M. D. Utidiian, and V. S. Karten. 1979. Retrospective evaluation of reproductive performance of workers exposed to ethylene dibromide (EDB). J. Occup. Med. 21:98-102. Woolson, E. A., R. F. Thomas, and P. D. J. Ensor. 1972. Survey of polychlorodibenzo- p-dioxin content in selected pesticides. J. Agric. Food Chem. 20(2):351-354. Wright, W. H., and J. M. Schaffer. 1932. Critical anthelmintic tests of chlorinated alkyl hydrocarbons and a correlation between the anthelmintic efficacy, chemical structure and physical properties. Am. J. Hyg. 16:325-428. Wyllie, J. A., J. Gabica, W. W. Benson, and J. Yoder. 1975. Exposure and contamination of the air and employees of a pentachlorophenol plant, Idaho 1972. Pestic. Monit. J. 9: 150-153. Wyrobek, A. J., and W. R. Bruce. 1975. Chemical induction of sperm abnollllalities in mice. Proc. Natl. Acad. Sci. USA 72:4425-4429. Yagi, Y., Y. Nakamura, I. Tomita, K. Tsuchikawa, and N. Shimoi. 1980. Teratogenic potential of di- and mono-(2-ethylhexyl)phthalate in mice. J. Environ. Pathol. Toxicol. 4(2,3):533-544. Yam, J., K. A. Booman, W. B;oddle, L. Geiger, J. E. Heinze, Y. J. Lin, K. McCarthy, S. Reiss, V. Sawin, R. I. Sedlak, R. S. Slesinski, and G. A. Wright. 1984. Surfactants: A survey of short-term genotoxicity testing. Food Chem. Toxicol. 22:761-769. Yoshikawa, K., A. Tanaka, T. Yamaha, and H. Kurata. 1983. Mutagenicity study of nine monoalkyl phthalates and a dialkyl phthalate using Salmonella typhimurium and Escherichia coli. Food Chem. Toxicol. 21:221-223. Zeiger, E., S. Haworth, W. Speck, and K. Mortelmans. 1982. Phthalate ester testing in the National Toxicology Program's environmental mutagenesis test development pro- gram. Environ. Health Perspect. 45:99-101.

Toxicity of Selected Contaminants 437 Zeiger, E., S. Haworth, K. Mortelmans, and W. Speck. 1985. Mutagenicity testing of di(2-ethyhexyl)phthalate and related chemicals in Salmonella. Environ. Mutagen. 7:213- 232. Zimmering, S. 1983. 1,2-Dibromo-3-chloropropane (DBCP) is positive for sex-linked recessive lethals, heritable translocations and chromosome loss in Drosophila. Mutat. Res. 119:287-288. Zober, A., K. H. Schaller, K. Gossler, and H. J. Krekeler. 1981. Pentachlorophenol and liver-function: A pilot-study on occupationally exposed collectives. Int. Arch. Occup. Environ. Health 48:347-356. (In German; English summary)