Guidance on Chloral Hydrate in Drinking Water

Part B - Supporting information

B.2 Health effects

B.2.3 Absorption, metabolism and excretion

In humans, chloral hydrate is rapidly absorbed and then either oxidized to TCA (8%) or reduced to TCOH (92%), mainly by the liver, but also by the kidney. TCOH may be conjugated with glucuronic acid to form trichloroethanol glucuronide (TCOG), an inactive metabolite (Ogino et al., 1990; McEvoy, 1999). Additional TCA is formed during enterohepatic circulation of TCOH, such that 35% of the initial dose of chloral hydrate is converted to TCA (Allen and Fisher, 1993). The erythrocytes also metabolize chloral hydrate to TCOH, mainly via alcohol dehydrogenase.

Healthy male volunteers (n = 18) were administered a single dose of 250 mg of chloral hydrate in drinking water. Chloral hydrate, TCOH, and TCA were measured in the plasma. Chloral hydrate could be detected in only some of the plasma samples, 8-60 minutes after dosing. No concentration was reported, but the limit of detection was given as 0.1 mg/L. The maximum plasma concentrations of TCOH and TCA, 3 mg/L and 8 mg/L, respectively, were achieved 0.67 hour and 32 hours after dosing, respectively. The terminal half-lives were 9.3-10.2 hours for TCOH and 89-94 hours for TCA (Zimmermann et al., 1998).

The plasma half-lives in humans for therapeutic doses of chloral hydrate, TCOH, and TCA are about 4-5 minutes, 8-12 hours, and 67 hours, respectively (Ellenhorn and Barceloux, 1988).

In infants and children given chloral hydrate as a sedative, TCA, DCA, and TCOH were detected in plasma, and TCA's plasma half-life was shown to be very long (Henderson et al., 1997). There is evidence that the TCA may be converted to DCA in samples of blood taken for analysis unless appropriate steps are taken, raising concerns about whether the reported levels of DCA in humans are too high (Ketcha et al., 1996).

Chloral hydrate and TCOH do not accumulate in the human body (Gilman et al., 1985). As infants have an immature hepatic metabolism, particularly the glucuronidation pathway, with decreased glomerular filtration, they have a longer TCOH half-life compared with adults. In contrast, toddlers have a TCOH half-life similar to that of adults, indicating maturation of liver metabolism in toddlers (IPCS, 2000).

Chloral hydrate is rapidly metabolized by rats and mice, producing both TCOH and TCA as the major metabolites, with a higher concentration of TCA in mice than in rats. The metabolism of chloral hydrate, TCA, and TCOH was shown in vitro to give rise to free radical intermediates that caused lipid peroxidation and the formation of malondialdehyde (Beland, 1999). As with humans, chloral hydrate disappears rapidly from the blood of mice, and TCOH, TCOG, TCA, and DCA are identified as metabolites (Abbas et al., 1996). The half-lives of TCOH and TCOG appear to be significantly greater in rats than in mice (Beland et al., 1998). Lipscomb et al. (1996) found TCOH to be the first major metabolite of chloral hydrate in vivo in the blood and liver of Fischer 344 rats, B6C3F1 mice, and humans.

Chloral hydrate is an important metabolite of trichloroethylene (TCE) and an intermediate in the formation of TCA. Based on the results from a number of studies, a physiologically based pharmacokinetic (PBPK) model for TCE was developed. This model includes enterohepatic recirculation of its metabolites. The model quantitatively predicts quite well the uptake, distribution, and elimination of TCE, TCOH, TCOG, and TCA. The PBPK model clearly shows that the formation of TCA is delayed following the enterohepatic recirculation, accounting for the longer half-life of TCA observed in animal studies (Stenner et al., 1998). This is supported by the findings of Merdink et al. (1999), who found that some TCOH is converted back to chloral hydrate and oxidized to TCA.

Most chloral hydrate is excreted via the urine as TCOG, with small amounts excreted as free TCOH. The remainder is excreted as TCA (Butler, 1948; Marshall and Owens, 1954; Allen and Fisher, 1993). Chloral hydrate is not excreted unchanged (McEvoy, 1999).

B.2.4 Risk assessment

No epidemiological or carcinogenic studies were found in humans that associated exposure to chloral hydrate with cancer, despite the fact that chloral hydrate has been used for many decades (and still is used) as a sedative and hypnotic drug in adults and children (specifically for dental procedures). The U.S. EPA (2000) derived an acute oral reference dose of 0.1 mg/kg bw per day based on the pharmacologically active dose (250 mg, equivalent to 10.7 mg/kg bw per day) in humans. This dose is said to be protective for any non-cancer health effects from chronic exposure. However, chloral hydrate has shown some evidence of carcinogenicity in two long-term drinking water bioassays in male mice and in a lifetime study following a single oral dose in male mice. In addition, chloral hydrate was found to be a weak mutagen and clastogen, suggesting that genotoxicity may play a role in the toxicity of chloral hydrate, but at concentrations higher than those expected to be found in the environment. The pharmacological dose of 10.7 mg/kg bw per day is not considered appropriate for the derivation of a health-based value for chloral hydrate in drinking water.

The International Agency for Research on Cancer classified chloral hydrate as Group 3, "not classifiable as to its carcinogenicity to humans," in 1995, based on inadequate evidence in humans and limited evidence in experimental animals (IARC, 1995). The U.S. EPA (2000) classified chloral hydrate as a possible human carcinogen, concluding that the most likely mode of action for the formation of tumours in mice involves interaction with cellular enzymes and proteins, in contrast to direct interaction with DNA. Health Canada has classified chloral hydrate in Group III.B -- possibly carcinogenic to humans (inadequate data in humans, limited data in animals), as defined in Health Canada (1994). There is equivocal evidence of genotoxicity for chloral hydrate.

For compounds that are "possibly carcinogenic to humans," a health-based value is based on a tolerable daily intake (TDI) derived by the division of the lowest NOAEL or LOAEL by appropriate uncertainty factors.

Two studies from the NTP (2002a, 2002b) provide weak evidence of carcinogenicity in B6C3F1 mice (both sexes). However, significant discrepancies exist between the experimental and historical control (Haseman et al., 1998) and dose group incidences of pituitary pars distalis adenomas and hepatocellular neoplasms and adenomas/carcinomas, making the studies unsuitable for derivation of a guideline. However, in these two studies, hepatocellular neoplasms developed at concentrations similar to those observed in the study chosen for the risk assessment (George et al., 2000), supporting the evidence of proliferative lesions at these concentrations.

The non-cancer end-point of histopathology in the liver as derived in George et al. (2000) was chosen for the risk assessment. Male B6C3F1 mice were treated in a lifetime study with chloral hydrate 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). 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. In this study, drinking water was used as a vehicle rather than gavage dosing 5 days per week as in the NTP (2002a, 2002b) studies, supporting the use of the George et al. (2000) study for this evaluation.

Although IPCS (2000) set a NOAEL of 1280 mg/L (146.6 mg/kg bw per day) for non-cancer end-points (based on the lack of evidence of hepatocellular necrosis at any exposure and only minimal changes in the levels of serum enzymes), the George et al. (2000) study showed that chloral hydrate induced an increase in the incidence of proliferative lesions (hyperplasia, adenoma, carcinoma, and combined adenoma and carcinoma) at all exposures, except for carcinoma at the two lower exposures. In the control groups, lesions (hyperplasia, adenoma, carcinoma, and combined adenoma and carcinoma) were also observed, but at lower percentages for the hyperplasia and hepatocellular adenoma. At 120 mg/L (13.5 mg/kg bw per day) and above, significant increases in the incidence of proliferative lesions were observed. This increase in proliferative lesions is an important end-point. Since these lesions were observed at all dose levels, no NOAEL could be derived; therefore, a LOAEL of 120 mg/L (13.5 mg/kg bw per day) was set to derive a TDI. An additional uncertainty factor of 3 was added to account for the limitations of the database in regards to evidence of carcinogenicity in animals.
The TDI is derived as follows:

where

• 13.5 mg/kg bw per day is the LOAEL for B6C3F1 mice based on increased incidence of liver histopathology,
• 3000 is the uncertainty factor (×10 for interspecies variability; ×10 for intraspecies variability; ×10 to account for the use of a LOAEL instead of a NOAEL, as no NOAEL was observed in the relevant study; and ×3 to account for limitations of the database in regards to evidence of carcinogenicity).

Based on this TDI, a health-based value for chloral hydrate in drinking water can be derived as follows:

where:

• 0.0045 mg/kg bw per day is the TDI, as derived above,
• 70 kg is the average body weight of an adult,
• 0.80 is the proportion of total daily intake allocated to drinking water, since the exposure to chloral hydrate for the general population comes mostly from chlorinated drinking water,
• 1.5 L/day is the average daily consumption of drinking water for an adult.

B.3 Detection and treatment

B.3.1 Analytical methods

The U.S. Environmental Protection Agency (EPA) recognizes and approves EPA Method 551.1 for the determination of chloral hydrate in drinking water. This method uses a solvent extraction procedure for the analysis of chloral hydrate, with methyl tert-butyl ether (MTBE) as the solvent (U.S. EPA, 1995). Chloral hydrate is analysed using gas chromatography/electron capture detection, and the method detection limit is 0.005 µg/L. The sampling protocol requires field pH adjustment (pH 4.8) with a phosphate buffer and use of sodium sulphite to quench the residual chlorine.

Standard Method 5710 D of the 21st edition of the Standard Methods for the Examination of Water and Wastewater is also used to analyse chloral hydrate (APHA et al., 2005). This method stipulates that chloral hydrate may be analysed with THMs using a sodium sulphite solution to quench the reaction. Chloral hydrate is then analysed using liquid-liquid extraction capillary column gas chromatography with electron capture detection.

B.3.2 Treatment technology

Although chloral hydrate formation in water is largely a function of the amount of organic compounds in water and their contact time with chlorine, it is important to recognize that the use of chlorination and other disinfection processes has virtually eliminated waterborne microbial diseases. As with THMs and other chlorinated DBPs, it is important to characterize the source water to ensure that the treatment process is optimized for precursor removal in order to reduce chloral hydrate levels in the finished water.

Municipal scale

The U.S. EPA has suggested that enhanced coagulation and softening will control chloral hydrate levels in drinking water by removing the DBP precursors (total organic carbon). Moving the point of disinfection to reduce the reaction between chlorine and DBP precursors and using chloramines instead of chlorine for residual disinfection have also been suggested as ways to reduce chloral hydrate production (U.S. EPA, 1998a). Controlling levels of total THMs and HAAs and using enhanced coagulation/softening for DBP precursor removal will control for chloral hydrate as well as other chlorination by-products (U.S. EPA, 1998b).

There are three approaches to limiting the concentrations of chloral hydrate in municipally treated drinking water:

• treatment of water to remove chloral hydrate precursors prior to disinfection;
• the use of alternative disinfectants and disinfection strategies; and
• treatment of water to remove chloral hydrate after its formation.

The majority of changes occurring in the water industry today focus on strategies to remove DBP precursors prior to disinfection and the use of alternative disinfectants and alternative disinfection strategies.

Removal of precursors prior to municipal disinfection

The removal of organic precursors is the most effective way to reduce the concentrations of all DBPs, including chloral hydrate, in finished water (U.S. EPA, 1999b; Reid Crowther & Partners Ltd., 2000). These precursors include synthetic organic compounds and natural organic matter, which can react with disinfectants to form chloral hydrate. Conventional municipal-scale water treatment techniques can reduce the amount of precursors, but are ineffective in removing chloral hydrate once it is formed. Granular activated carbon, membranes, and ozone biofiltration systems can also remove organic matter from water. The U.S. EPA has identified precursor removal technologies such as granular activated carbon (GAC) and membrane filtration as Best Available Technologies (BAT) for controlling disinfection by-products formation (U.S. EPA, 2005). However, membrane processes generate concentrated residuals, and their disposal can be expensive (Xie, 2004). Combinations of disinfectants, when optimized, can help control chloral hydrate formation.

Potassium permanganate can be used to oxidize organic precursors at the head of the treatment plant, thus minimizing the formation of DBPs at the disinfection stage (U.S. EPA, 1999a). The use of ozone for oxidation of precursors is currently being studied. Early work has shown that the effects of ozonation, prior to chlorination, depend on treatment design and raw water quality and thus are unpredictable. The key variables that seem to determine the effect of ozone are dose, pH, alkalinity, and the nature of the organic material in the water. Ozone has been shown to be effective at reducing precursors at low pH. Above pH 7.5, however, ozone may actually increase the production of chlorinated DBP precursors (U.S. EPA, 1999a).

Alternative municipal disinfection strategies

The use of alternative disinfectants, such as chloramines (secondary disinfection only), ozone (primary disinfection only), and chlorine dioxide (primary disinfection only), is increasing. However, each of these alternatives has also been shown to form its own set of DBPs. Preozonation is feasible for water sources that have turbidity levels below 10 nephelometric turbidity units and bromide concentrations below 0.01 mg/L, to minimize the formation of bromate (Reid Crowther & Partners Ltd., 2000). Ultraviolet (UV) disinfection is also being used as an alternative disinfectant. Since UV disinfection is dependent on light transmission to the microbes, water quality characteristics affecting UV transmittance must be considered in the design of the system. Neither ozone nor UV disinfection leaves a residual disinfectant, and both must therefore be used in combination with a secondary disinfectant to maintain a residual in the distribution system.

It is recommended that any change made to the treatment process, particularly when changing the disinfectant, be accompanied by close monitoring of lead levels in the distributed water. A change of disinfectant has been found to affect the levels of lead at the tap, for example in Washington, DC, where a change from chlorine to chloramines resulted in significantly increased levels of lead in the distributed drinking water. When chlorine, a powerful oxidant, is used as the disinfectant, lead dioxide scales formed in distribution system pipes have reached a dynamic equilibrium in the distribution system. In Washington, DC, switching from chlorine to chloramines decreased the oxidation-reduction potential of the distributed water and destabilized the lead dioxide scales, which resulted in increased lead leaching (Schock and Giani, 2004). Subsequent laboratory experiments by Edwards and Dudi (2004) and Lytle and Schock (2005) confirmed that lead dioxide deposits could be readily formed and subsequently destabilized in weeks to months under realistic conditions of distribution system pH, oxidation-reduction potential and alkalinity.

Residential scale

Municipal treatment of drinking water is designed to reduce contaminants to levels at or below their guideline values. As a result, the use of residential-scale treatment devices on municipally treated water is generally not necessary but primarily based on individual choice. In cases where municipal treatment has produced low concentrations of chloral hydrate in drinking water, some residential-scale point-of-entry or point-of-use treatment devices may remove chloral hydrate from the water. Treatment device technologies that may remove chloral hydrate include reverse osmosis and adsorption media, such as activated carbon, although none is currently certified specifically for this use.

NSF International (NSF) has developed several standards for residential water treatment devices designed to reduce the concentrations of various types of contaminants in drinking water, but chloral hydrate is not currently included in any NSF standard. Research is ongoing in the private and public sectors to test and adopt efficient methods for the reduction of chloral hydrate levels in drinking water.

Devices can lose removal capacity through usage and time and need to be maintained and/or replaced. Consumers should verify the expected longevity of the adsorption media in their treatment devices as per the manufacturers' recommendations and service the media when required.

Health Canada has conducted a study on the effectiveness of a number of point-of-use drinking water treatment devices for the removal of chloral hydrate. Boiling water for 2-5 minutes in a pot or kettle was effective in removing chloral hydrate (~98% decrease). The efficiency of filters (pressure filters and gravity filters using granular activated carbon) for the removal of chloral hydrate largely depended on the brand and age of the filters (new filters, 28 to >99% decrease). Aging of the filter, even in the short term, significantly reduced its capacity to remove chloral hydrate (Benoit et al., 2000; LeBel et al., 2002).

Health Canada does not recommend specific brands of drinking water treatment devices, but it strongly recommends that consumers look for a mark or label indicating that the device has been certified by an accredited certification body to the appropriate NSF/American National Standards Institute (ANSI) drinking water materials standard. These standards have been designed to safeguard drinking water by helping to ensure the material safety and performance of products that come into contact with drinking water. Certification organizations provide assurance that a product conforms to applicable standards and must be accredited by the Standards Council of Canada (SCC). In Canada, the following organizations have been accredited by the SCC to certify drinking water devices and materials as meeting NSF/ANSI standards:

•  Canadian Standards Association International (www.csa-international.org);
•  NSF International  (www.nsf.org);
•  Water Quality Association  (www.wqa.org);
•  Underwriters Laboratories  (www.ul.com);
•  Quality Auditing Institute  (www.qai.org); and
•  International Association of Plumbing and Mechanical Officials  (www.iapmo.org). An up-to-date list of accredited certification organizations can be obtained from the SCC  (www.scc.ca).