Guidance on Chloral Hydrate in Drinking Water

Part B - Supporting information

B.1 Physical/environmental considerations

B.1.1 Identity, use, sources, and fate in the environment

Chloral hydrate (2,2,2-trichloro-1,1-ethanediol) has a relative molecular mass of 165.4, a crystalline appearance, an aromatic and slightly acrid odour, and a slightly bitter taste (Reynolds and Prasad, 1982; Budavari, 1996). It is synthesized by the chlorination of ethanol (Reynolds and Prasad, 1982; Budavari, 1996; Verschueren, 2001).

Chloral hydrate has a melting point of 57°C, a boiling point of 96°C (Hansch et al., 1995), and a density/specific gravity of 1.91 g/cm³ at 20°C. It has an octanol/water partition coefficient (log K_ow) of 0.99; therefore, bioconcentration is not an important factor. At 25°C, chloral hydrate has a vapour pressure of 2 kPa (Reynolds and Prasad, 1982; Hansch et al., 1995) and a water solubility of 9.3 × 10⁶ mg/L (McEvoy, 1999). It slowly volatilizes when exposed to ambient air, and it decomposes when exposed to light (McEvoy, 1999).

Chloral hydrate is used as a sedative and hypnotic in human and veterinary medicine. It is also used in the manufacture of DDT (Budavari, 1996) and dichloroacetic acid (DCA) (Kirk-Othmer, 1991). In addition, chloral hydrate is used as an intermediate in the production of the insecticides methoxychlor, naled, trichlorfon, and dichlorvos, the herbicide trichloroacetic acid (TCA), and the hypnotic drugs chloral betaine, chloralose, and trichlorfos sodium (IARC, 1995). Chloral hydrate can be formed as a by-product of the chlorination of water containing organic precursor molecules, such as fulvic and humic acids. Chloral hydrate can also be released to the environment from wastewater treatment facilities, from the manufacture of pharmaceutical-grade chloral hydrate, and from the waste stream during the manufacture of insecticides and herbicides that use chloral hydrate as an intermediate (U.S. EPA, 2000).

Chloral hydrate can be transformed into trichloroethanol (TCOH) and TCA by the methanotrophic bacterium Methylosinus trichosporium. The transformation of chloral hydrate into chloroform occurs under abiotic conditions (pH 9.0 and 60°C) after 2.3 minutes. Formic acid is another decomposition product of chloral hydrate (Newman and Wackett, 1991).

B.1.2 Exposure

No data are available on human exposure to chloral hydrate in air. The high water solubility and low volatility of chloral hydrate preclude significant exposure by inhalation from a water solution (U.S. EPA, 2000).

According to surveys conducted in Canada in 1995 and 1997, the mean level of chloral hydrate in drinking water ranged from 1.2 to 3.8 µg/L in winter and from 3.6 to 8.4 µg/L in summer, with a maximum level of 22.5 µg/L observed in winter from a sampling of 53 sites (Health Canada, 1995; Edsall and Charlton, 1997; Williams et al., 1997). Although slightly higher levels of chloral hydrate may be found in smaller treatment systems with limited ability to remove organic matter before adding the chlorine disinfectant, levels are still below any level of concern.

In the United States, median chloral hydrate concentrations in finished water have been reported to range from 1.7 to 2.5 µg/L, whereas maximum concentrations ranged from 22 to 46 µg/L (Krasner et al., 1989; U.S. EPA, 1992). The chloral hydrate concentration was higher in distribution systems of surface water plants (median 4.0 µg/L) than in groundwater (median 0.5 µg/L) and was generally higher than the concentration in the finished water (median 2.4 µg/L), suggesting that chloral hydrate concentrations increase across the distribution system.

No data are available on human exposure to chloral hydrate in food (IARC, 1995).

For adults, the usual hypnotic dose of chloral hydrate is 0.5-1 g, whereas the usual sedative dosage is 250 mg 3 times daily. When chloral hydrate is administered in the management of alcohol withdrawal symptoms, the usual dosage is 0.5-1 g repeated at 6-hour intervals if needed. Generally, single doses or daily dosages for adults should not exceed 2 g. For children, the hypnotic dose of chloral hydrate is 50 mg/kg bw or 1.5 g/m³, with a maximum dose of 1 g. The sedative dosage for children is 8 mg/kg bw or 250 mg/m³ 3 times daily, with a maximum dosage of 500 mg 3 times a day. As a premedication before electroencephalogram evaluation, children have been given chloral hydrate at a dose of 2-25 mg/kg bw (McEvoy, 1999).

Chloral hydrate is highly water soluble, has a log K_ow of less than 10, and is known to occur in drinking water supplies as a DBP. There are no data on the levels of chloral hydrate in air, soil and food, and there is no indication that it would be present at significant levels in these environmental media. These characteristics suggest that drinking water would be the primary source of exposure to chloral hydrate for the general population; therefore, an allocation factor of 80% is used in the risk assessment. Occupational exposure during manufacturing (IARC, 1995) and exposure from pharmaceutical use of chloral hydrate may also occur.

B.2 Health effects

B.2.1 Effects in humans

Chloral hydrate was introduced into therapeutics more than 100 years ago and has been used as a sedative/hypnotic agent in children, adults, and animals since its introduction (Henderson et al., 1997). Insufficient data are available to determine a no-observed-adverse-effect level (NOAEL) in humans. The lowest-observed-adverse-effect level (LOAEL) is 10.7 mg/kg bw per day (assuming a body weight of 70 kg), based on the recommended dose as a sedative for an adult of 250 mg 3 times a day.

Oral administration of chloral hydrate at high doses causes gastric irritation, with nausea, vomiting, and diarrhoea as the most frequent adverse effects. Other adverse effects of chloral hydrate may include leukopenia, eosinophilia, and, rarely, ketonuria (McEvoy, 1999).

The toxic blood level and the lethal blood level for chloral hydrate were estimated to be 10 mg/100 mL and 25 mg/100 mL, respectively (Ellenhorn et al., 1997).

While a lethal oral dose of 10 g has been reported for adults, death has occurred after ingestion of 4 g, and some patients have survived ingestion of as much as 30 g (McEvoy, 1999). Ingestion of 20 g by a patient, who later became comatose, resulted in gastric perforation that was detected 4 days post-ingestion. Gastrointestinal haemorrhage followed by the development of oesophageal strictures has been observed with a dose of 18 g. Hepatic (jaundice, aminotransferase elevation) and renal (albuminuria) dysfunction may occur several days after ingestion, but is rarely serious or prolonged (Abbas et al., 1996).

A variety of adverse effects were reported in 1618 patients who had received chloral hydrate at various doses, although it was not clear if the patients had the identified clinical effects prior to being exposed to chloral hydrate or if they developed the clinical effects after being exposed to chloral hydrate. The results indicated that cirrhosis of the liver was the most common diagnosis (15%), whereas chronic obstructive respiratory tract disease (7%), carcinoma of the breast (7%), and congestive cardiac failure (7%) were also reported, although a causal association could not be determined. Other adverse reactions, including gastrointestinal symptoms (10 patients), depression of the central nervous system (20 patients), skin rash (5 patients), prolonged prothrombin time (1 patient), worsened hepatic encephalopathy (1 patient), and bradycardia (1 patient), disappeared soon after the end of chloral hydrate administration (Shapiro et al., 1969). Another review of medical records has shown central nervous system depression to be the preponderant effect following exposure to chloral hydrate in 5435 patients (Greenberg et al., 1991).

No long-term studies of chloral hydrate exposure in humans were available in the published literature.

B.2.2 Effects on experimental animals and in vitro

Acute toxicity

The LD₅₀ for chloral hydrate in mice was determined to be 1265 mg/kg bw for females and 1442 mg/kg bw for males. Rats were found to be more sensitive to chloral hydrate, with LD₅₀s of 285 and 479 mg/kg bw for newborn pups and adults, respectively (Sanders et al., 1982).

Mice exposed to chloral hydrate at 603 mg/m³ for 6 hours by inhalation exhibited several changes in the lung, including vacuolization of Clara cells, alveolar necrosis, desquamation of the epithelium, and alveolar oedema (Odum et al., 1992).

Short-term exposure

The liver is the primary target organ of short-term exposure to chloral hydrate. In a 7-day study with 28 male Sprague-Dawley rats administered chloral hydrate in drinking water at dose levels of 5, 43, or 375 mg/kg bw per day, no NOAEL could be determined, as no histopathological changes were observed in the liver. However, other changes in the liver (e.g., increase in hepatic peroxisomal enzyme palmitoyl coenzyme A [CoA] oxidase, suppression of hepatic aldehyde dehydrogenase [ALDH] activity, decreases in liver cholesterol and liver triglyceride levels) suggested that the liver is the target organ of chloral hydrate exposure (Poon et al., 2000). In a 13-week study by the same investigators in which Sprague-Dawley rats (10 per sex per dose) were administered chloral hydrate in drinking water at 0, 0.2, 2, 20, or 200 mg/L (corresponding to dose levels of 0, 0.02, 0.19, 1.9, and 19.8 mg/kg bw per day for males and 0, 0.03, 0.24, 2.6, and 23.6 mg/kg bw per day for females), the no-observed-effect level (NOEL) was identified as 1.9 mg/kg bw per day in males and 2.6 mg/kg bw per day in females based on the decrease in ALDH activity in both sexes at the highest dose, the increase in aniline hydroxylase activity in both sexes at the highest dose, and the minimal vacuolation of the myelin sheath in males at the highest dose. The LOAEL for males in this study was 19.8 mg/kg bw per day, based on the mild vacuolation of the myelin sheath (Poon et al., 2002). (The authors stated that nervous tissue is particularly susceptible to inadequate fixation, with vacuolation being one of the most common histological artefacts.)

In a study in which CD-1 mice (48 per sex for control; 32 per sex for treatment groups) were administered chloral hydrate in drinking water for 90 days at concentrations of 70 or 700 mg/L (corresponding to dose levels of 16 and 160 mg/kg bw per day for males and 18 and 173 mg/kg bw per day for females), a LOAEL of 160 mg/kg bw per day and a NOAEL of 16 mg/kg bw per day were identified, based on changes observed in the liver of males, including increased liver weight, hepatomegaly, and microsome proliferation (Sanders et al., 1982). In a similar study, the authors determined a NOAEL of 16 mg/kg bw per day for decreased humoral immunity (assessed by verifying the number of splenic antibody-forming cells produced in response to sheep red blood cells and haemagglutination titres) and a LOAEL of 160 mg/kg bw per day (Kauffmann et al., 1982).

The liver was confirmed as the target organ in a study in which Sprague-Dawley rats (10 per sex per dose) were exposed to chloral hydrate for 90 days in drinking water at concentrations of 300, 600, 1200, or 2400 mg/L (corresponding to dose levels of 24, 48, 96, and 168 mg/kg bw per day for males and 33, 72, 132, and 288 mg/kg bw per day for females). Based on hepatotoxic effects (focal hepatocellular necrosis in males) and serum enzyme changes (observed in both sexes, but not dose related or toxicologically significant), the study identified a LOAEL of 96 mg/kg bw per day and a NOAEL of 48 mg/kg bw per day (Daniel et al., 1992b).

Long-term exposure and carcinogenicity

The liver has been confirmed as the primary target organ of chloral hydrate toxicity in several long-term and carcinogenicity studies.

In a drinking-water study in mice (Rijhsinghani et al., 1986), chloral hydrate at 5 or 10 mg/kg bw was given to 15-day-old male C57BL × C3HF1 mice as a single dose in distilled water. The control group was given distilled water only. To study long-term effects, animals were sacrificed when found moribund or at intervals up to 92 weeks. An increase in tumours was statistically significant (P < 0.05) only in the 10 mg/kg bw group. There was an increase in the relative weight of the liver in mice given chloral hydrate at 10 mg/kg bw, but not in mice given 5 mg/kg bw. Compared with the control group, there was also a significant increase in the incidence of hepatic nodules in mice given chloral hydrate at a dose of 10 mg/kg bw only. The hepatic nodular lesions ranged from hyperplastic foci of clear or acidophilic cells to hepato-cellular adenomas and trabecular carcinomas containing eosinophilic hepatocytes (Rijhsinghani et al., 1986). It is important to note that this study is more than 20 years old and that the protocol used was not based on Organisation for Economic Co-operation and Development (OECD) guidelines regarding the use of two sexes, a required number of animals, and a sufficient number of doses for a carcinogenicity study. The increased incidence of hepatic tumours in this study is believed to have been due to normal variation in mice and not a result of chloral hydrate treatment (NTP, 2002b).

In a key study, male B6C3F1 mice (72 per dose) were exposed to chloral hydrate in drinking water at concentrations of 0, 120, 580, or 1280 mg/L (corresponding to dose levels of 0, 13.5, 65, and 146.6 mg/kg bw per day) for 104 weeks (George et al., 2000). The prevalence of hepatocellular carcinomas was increased in the high-dose group (84.4%) compared with 54.8%, 54.3%, and 59.0% in the control, low-, and mid-dose groups. The prevalence of hepatoadenomas was significantly increased in all dose groups: 43.5%, 51.3%, and 50.0% for the low-, mid-, and high-dose chloral hydrate groups, respectively, compared with 21.4% in the control group. Serum lactate dehydrogenase (LDH), alanine aminotransferase (ALT), aspartate aminotransferase (AST), and sorbitol dehydrogenase (SDH) activities and total antioxidant levels reflected the minimal degree of hepatocellular damage observed microscopically. None of these parameters in the chloral hydrate-treated groups was altered compared with the control values after 52 and 78 weeks of exposure. Palmitoyl CoA oxidase activities in the homogenates of livers were not significantly increased above the control value, indicating that chloral hydrate did not induce peroxisome proliferation. Enhanced liver neoplasia occurred at the lowest dose, 13.5 mg/kg bw per day; therefore, a NOAEL could not be determined. However, a LOAEL can be set at 13.5 mg/kg bw per day. Results for combined neoplasms were significantly increased in the mid- and high-dose groups for prevalence and in all dose groups for multiplicity. This study indicated that the incidence of hepatocellular adenomas was increased at all dose levels, but the incidence of hepatocellular carcinomas was increased at the high dose only (George et al., 2000). IPCS (2000) evaluated this study and set a NOAEL of 146.6 mg/kg bw per day for non-cancer effects in mice, justifying a NOAEL for non-cancer end-points because of the prevalence of proliferative lesions in the controls. It was also noted that there was no increase in the prevalence of neoplasia at sites other than the liver. The male mice showed an increase of proliferative lesions in the liver at all exposure levels (hyperplasia, adenoma, carcinoma, and combined adenoma and carcinoma).

In another component of the study (George et al., 2000), male F344 rats (78 per dose) were administered chloral hydrate in drinking water at concentrations of 0, 120, 580, or 2510 mg/L (corresponding to dose levels of 0, 7.4, 37.4, and 162.6 mg/kg bw per day) for 104 weeks. No measured parameters in any of the chloral hydrate-treated groups were altered compared with the control values after 52 and 78 weeks of exposure, and the NOAEL for this study was set at the highest dose (George et al., 2000). However, the U.S. EPA (2000) concluded that this study did not achieve the maximum tolerated dose, and we concur with this conclusion.

Two National Toxicology Program (NTP) studies (NTP, 2002a, 2002b) provide weak evidence of the carcinogenicity of chloral hydrate in male and female B6C3F1 mice. In the first study (NTP, 2002a), five groups of B6C3F1 mice received chloral hydrate in distilled water by gavage for 105 weeks in a complicated dosing regimen. Chloral hydrate doses ranging up to 100 mg/kg bw per day were administered 5 days per week (71.4 mg/kg bw per day adjusted for 7 days per week dosing). In the first group, the incidence of pituitary pars distalis adenomas occurred with a positive dose-related trend, and the incidence at the highest dose was significantly greater than that in controls; there was also a significant positive time-related trend in the incidence of adenomas and a significant increase in the severity of pars distalis hyperplasia in female mice administered 71.4 mg/kg bw per day for up to 24 months. The authors concluded that there was equivocal evidence of carcinogenic activity of chloral hydrate in female mice treated continuously for 2 years, based on increased incidences of pituitary gland pars distalis adenomas. No increased incidences of neoplasms were seen in female B6C3F1 mice that received a single dose of chloral hydrate at 15 or 28 days of age or in male B6C3F1 mice that received a single dose of chloral hydrate at 15 days of age. No hepatic carcinogenicity was seen under all dosing conditions. The NOAEL for non-neoplastic effects was determined to be 71.4 mg/kg bw per day (NTP, 2002a).

Haseman et al. (1998) reported significant discrepancies between the experimental and historical control in the NTP (2002a) study, as well as dose group incidences of pituitary pars distalis adenomas and hepatocellular neoplasms and adenomas/carcinomas. For example, the incidence of adenomas (12% at 71.4 mg/kg bw per day) exceeded the incidence in historical controls (historical incidence for control groups in National Center for Toxicological Research [NCTR] studies: 4.9%, range 0-6%). The incidence of pituitary pars distalis adenomas/ carcinomas reached 14.8% in the historical control (Haseman et al., 1998), which is higher than the incidence observed at the highest dose (12%) in the first two test groups. This type of adenoma was not observed in any other studies. In both cases (pituitary gland adenoma and hyperplasia), no dose-related effects can be inferred. The incidence of malignant lymphomas found at the low and high doses (33% at 10 mg/kg bw) (in the third group) is higher than the incidence found in both control groups for the NCTR (24.6%) and Haseman et al. (1998) (20.9%) studies. The incidences of malignant lymphomas were not affected by the duration of the treatment and were within the range of incidences in the historical controls, inferring that the incidences observed in the first and third test groups reflect the normal variation found in female B6C3F1 mice. The incidence of alveolar/bronchiolar adenomas was significantly higher (18%) in the second group after 12 months. However, no significant difference was observed after 2 years, suggesting that the incidence may not have been due to chloral hydrate. The incidence found in this study is within the historical range (5.9%, range 0-24%) for control female B6C3F1 mice fed NIH-07 (an open formula cereal-based diet, the NTP standard for chemical toxicity and carcinogenicity studies) (Haseman et al., 1998), but it is higher than that observed in the other NCTR study in female B6C3F1 mice (3.8%, range 2-6%). No other studies have reported this type of lesion with chloral hydrate. In the NTP (2002a) study, the incidence of hepatocellular neoplasms in females in the third and fourth test groups was quite low, ranging from 1% to 13% at up to 50 mg/kg bw; in males (in the fifth group), the incidence of hepatocellular neoplasms was quite high, even in the control group (50%). There were no neoplasms in mice that could be attributed to chloral hydrate treatment.

In the second NTP study (NTP, 2002b), a group of 120 male B6C3F1 mice was fed 0, 25, 50, or 100 mg chloral hydrate/kg bw per day, 5 days per week (0, 17.9, 35.7, or 71.4 mg/kg bw per day, adjusted for 7 days per week dosing), for 2 years. The male mice were divided into two groups of 60: one group received feed ad libitum, while the other received feed in a measured daily amount (gavage). Evaluation of 12 mice per dose group and diet was performed after 15 months, and the other 48 mice per dose group and diet were evaluated after 2 years. Histopathological changes were observed only in the liver. The incidence of hepatocellular adenomas or carcinomas was significantly greater in the 17.9 mg/kg bw per day group only in the ad libitum group. In the dietary controlled study, the incidence of hepatocellular carcinomas was significantly different at 71.4 mg/kg bw per day only. The NOAEL for non-neoplastic effects is 71.4 mg/kg bw per day (NTP, 2002b). The incidences of hepatocellular adenomas/ carcinomas for the ad libitum group in male B6C3F1 mice were 33%, 52%, 48%, and 46% at dose levels of 0, 17.9, 35.7, and 71.4 mg/kg bw per day, respectively, compared with 42% in the historical controls. In the controlled diet, the incidences of combined adenomas and carcinomas were lower than the historical control of 42.2% with the same strain of mice, at 23%, 23%, 29%, and 38% for dose levels of 0, 17.9, 35.7, and 71.4 mg/kg bw per day, respectively. No female mice were treated in this study.

In a study by Daniel et al. (1992a), 40 male B6C3F1 mice received 1 g chloral hydrate/L (166 mg/kg bw per day) in drinking water for 104 weeks. Only mild histopathological changes were observed in the liver, and no changes were noted in other organs. At the 60-week sacrifice, hepatocellular carcinomas were found in two chloral hydrate-treated mice, compared with zero in the controls. No lesions were found in the control group at this time. At the end of the study, 11 out of 24 surviving mice exposed to chloral hydrate had hepatocellular carcinomas, and 7 had hepatocellular adenomas. Hepatocellular lesions in controls included 2 animals with carcinomas and 1 with adenomas out of 20 survivors examined. The incidence of both adenomas and carcinomas was significantly greater than in the control group. The increase in the combined incidence of the two lesions, adenomas and carcinomas, was highly significant (Daniel et al., 1992a). In this study, a low incidence of hepatocellular tumours was observed in the control group (15%), compared with 42.2% for the historical control (Haseman et al., 1998). However, the study indicated that the liver is the target organ following exposure to chloral hydrate. Hepatocellular necrosis as well as tumours (carcinoma, adenoma) were observed in this study. However, this study would be considered inadequate based on OECD guideline standard protocols, because only one dose was used. The difference in the incidences of hepatocellular adenomas and carcinomas in the NTP (2002a) study and that by Daniel et al. (1992a) might be due to the higher dose used in the latter study.

A chronic bioassay was conducted with Sprague-Dawley rats (50 per sex per group) administered chloral hydrate at doses of 0 (untreated drinking water), 15, 45, or 135 mg/kg bw per day in drinking water for 124 weeks for males and 128 weeks for females. There was no evidence of an increased incidence of tumours in any organs. An increase in the incidence of hepatocellular hypertrophy was observed at the highest dose (28%) compared with the control (11%). The finding was characterized by diffuse liver cell enlargement with slightly eosinophilic cytoplasm. The increase in hepatocellular hypertrophy was seen only in the male rats and was graded as minimal to slight in severity. The type, incidence, and organ distribution of the neo-plastic lesions in the chloral hydrate-treated rats did not differ from those in the control rats, and the lesions were therefore regarded as random events. No change was observed in body weight or organ weight. A NOAEL of 45 mg/kg bw per day and a LOAEL of 135 mg/kg bw per day were set based on the increased incidence of hepatocellular hypertrophy (Leuschner and Beuschner, 1998). The U.S. EPA (2000) concluded that there are indications that the study did not achieve the maximum tolerated dose, and we concur with that conclusion.

Mutagenicity and genotoxicity

There is equivocal evidence of the genotoxicity of chloral hydrate. Moore and Harrington-Brock (2000) found chloral hydrate and its metabolites to show evidence of some genotoxic activity, albeit at very high doses, indicating that chloral hydrate is a weakly genotoxic chemical.

In in vitro tests, positive results were reported for Salmonella typhimurium strains TA198 and TA100 in point mutation assays (Waskell, 1978; Bruce and Heddle, 1979; Bignami et al., 1980; NTP, 2002a), for sister chromatid exchanges (Bruce and Heddle, 1979; NTP, 2002a), for DNA strand breaks in human lymphocytes (Gu et al., 1981), for chromosomal aberrations in Chinese hamster ovary cells with or without S9 (Bruce and Heddle, 1979; NTP, 2002a), for aneuploidy and clastogenicity in several test systems using mammalian cells (Natarajan et al., 1993; Harrington-Brock et al., 1998), and for aneuploidy in various fungi in the absence of metabolic activation (U.S. EPA, 2000). Chloral hydrate induced aneuploidy in Chinese hamster embryonic fibroblasts in vitro without an exogenous metabolic system at 250 µg/mL (Harrington-Brock et al., 1998), in two Chinese hamster pulmonary lines at 250 µg/mL and 50 µg/mL (Warr et al., 1993), and in human peripheral blood lymphocytes at 50 µg/mL without an exogenous metabolic system (Sbrana et al., 1993).

Increases in micronuclei in mouse spermatids were observed when spermatogonia stem cells were exposed to chloral hydrate at 41, 83, or 165 mg/kg bw (Allen et al., 1994). In Drosophila melanogaster, chloral hydrate induced a small increase in the frequency of sex-linked recessive lethal mutations in germ cells of male flies fed 5500 mg/kg bw (Yoon et al., 1985).

Negative results were reported for S. typhimurium strain TA1535 in point mutation assays, for mitotic crossing-over in Aspergillus nidulans in the absence of metabolic activation, for reverse mutation in Saccharomyces cerevisiae, for DNA-protein crosslinks in rat liver nuclei, for DNA single strand breaks in rodent hepatocytes in primary culture and in CCRFCEM cells, a human lymphoblastic leukaemia cell line (Chang et al., 1992), for DNA repair in Escherichia coli (U.S. EPA, 2000), and for micronuclei induction and aneuploidy in mouse lymphoma cells (Natarajan et al., 1993; Harrington-Brock et al., 1998). Chloral hydrate also gave negative test results in studies with ICR mouse metaphase II oocytes (Mailhes et al., 1993).

In vivo, chloral hydrate was positive in the mouse bone marrow micronucleus test (U.S. EPA, 2000; NTP, 2002b). Chloral hydrate increased the frequency of chromosomal aberrations in mouse bone marrow, spermatogonia, and primary and secondary spermatocytes, but not in oocytes, after in vivo treatment (U.S. EPA, 2000).

Reproductive and developmental toxicity

The reproductive, embryo-fetotoxic, and teratogenic effects of chloral hydrate have been studied in several species. Effects have included neurodevelopmental toxicity, reduction in sperm motility, and an increased incidence of heart malformations.

Male and female CD-1 mice (four per cage, total number not specified) were exposed to chloral hydrate in drinking water at concentrations of 60 or 600 mg/L (equivalent to 21.3 or 204.8 mg/kg bw per day). All animals were exposed for 3 weeks prior to breeding. Females were also exposed during gestation and until pups were weaned at 21 days of age. At 204.8 mg/kg bw per day, at 23 days of age, pups showed impaired retention of passive avoidance learning in both 1-hour and 24-hour retention tests. This study identified a NOAEL for neurodevelopmental toxicity of 21.3 mg/kg bw per day. A reproductive and developmental effects NOAEL was also identified at 204.8 mg/kg bw per day (Kallman et al., 1984).

Male F344 rats (two per cage, total number not specified) were administered chloral hydrate in drinking water at concentrations of 0, 780, or 2700 mg/L (corresponding to dose levels of 0, 55, and 188 mg/kg bw per day) for 52 weeks to evaluate the effects of chloral hydrate on sperm morphology and motility. A reduction in sperm motility was observed in rats exposed to 188 mg/kg bw per day (58%) compared with controls (68%). A shift in the frequency distribution of the average straight-line velocities of sperm also occurred at this dose compared with the controls. A NOAEL for effects on sperm motility was set at 55 mg/kg bw per day, and a LOAEL of 188 mg/kg bw per day was identified (Klinefelter et al., 1995).

An in vitro embryotoxicity study was performed with embryos from Sprague-Dawley rats exposed to chloral hydrate at 0.5-2.5 mmol/L (83-414 mg/L) on gestational day 10 for 46 hours. All embryos died at the highest dose, but no deaths were observed at lower doses. Chloral hydrate caused concentration-dependent decreases in growth and differentiation and increases in the incidence of morphologically abnormal embryos. At 1.0 mmol/L, decreases in crown-rump length, somite (embryonic segments) numbers, and the protein or DNA content of embryos were observed. At 1.0, 1.5, and 2.0 mmol/L, 18%, 68%, and 100% of embryos, respectively, had malformations, malformations of the brain, eyes, and ears being among the most frequent developmental effects encountered. At 2.0 mmol/L, abnormalities were also observed in the trunk and in the optic and otic systems. At the higher concentrations, embryos exhibited severe alterations of the craniofacial region. Hypoplasia of the prosencephalon was also observed. Chloral hydrate caused pericardial dilation (45% of the embryos at 2 mmol/L). Chloral hydrate produced a step increase in embryolethality as concentrations increased. Based on this in vitro study, chloral hydrate was found to be more potent than TCA and DCA. A NOAEL of 0.5 mmol/L (83 mg/L) was identified for embryotoxicity (Saillenfait et al., 1995). Since this study is an in vitro study, it is difficult to compare the results with those of an in vivo study, nor is it possible to extrapolate the risk to human health. This study is unusual, because reproductive studies tend to involve exposure over the period of organogenesis (days 5-12) or over the entire gestation period.

Pregnant female Sprague-Dawley rats were exposed to chloral hydrate in drinking water from gestational day 1 to day 22 at 1.232 mg/mL (corresponding to an average exposure of 151 mg/kg bw per day). There was no evidence of maternal toxicity, no change in the number of live or dead fetuses, no change in placental or fetal weight, no change in crown-rump length, and no increase in the incidence of morphological changes. Heart malformations (not significant), such as atrial septal defects (2), mitral valve defects (2), ventricular septal defects (3), and pulmonary valve defects (1), were observed, compared with the control group, which had a total of 15 types of heart malformations. The chloral hydrate-exposed group had a 3.23% incidence of heart abnormalities compared with 2.15% for the control group (significantly different). A LOAEL of 151 mg/kg bw per day for chloral hydrate was identified in this study based on developmental toxicity, and no NOAEL was identified (Johnson et al., 1998).

Mode of action

In the Poon et al. (2002) subchronic toxicity study described above, it was postulated that the biological effects observed were attributable to TCA, a known peroxisomal proliferator. Triglyceride depression may also be a TCA effect. The presence of TCA in the serum and increased liver catalase lend support to a peroxisomal proliferation effect of chloral hydrate. However, this is of limited relevance in humans, since humans and other primates are less responsive than rats and mice in terms of peroxisomal proliferation. In contrast, hepatic hypotriglyceridaemia is of relevance to humans, because the hypolipidaemic effect of peroxisome proliferators is common to both experimental animals and humans.

In an in vitro study using male B6C3F1 mouse liver microsomes, it was found that chloral hydrate generated free radical intermediates that resulted in endogenous lipid peroxidation, thus forming malondialdehyde, formaldehyde, acetaldehyde, acetone, and propionaldehyde, substances that are known to be tumorigenic. Both TCA and TCOH also induced lipid peroxidation, but TCA had a stronger effect than TCOH, suggesting that TCA formation is the predominant pathway leading to lipid peroxidation. Cytochrome P-450 (possibly the isoenzyme CYP2E1) was the enzyme responsible for the metabolic activation of chloral hydrate and its metabolites to TCA and TCOH, leading to tumorigenic lipid peroxidation (Ni et al., 1996).