Page 11: Guidelines for Canadian Drinking Water Quality: Guideline Technical Document – Tetrachloroethylene

• Previous Page
• Table of Contents
• Next Page

10.0 Classification and assessment

Tetrachloroethylene was classified as a Group III carcinogen (possibly carcinogenic to humans) in the Priority Substances List Assessment Report (Government of Canada, 1993). The International Agency for Research on Cancer (IARC) recently classified tetrachloroethylene as probably carcinogenic to humans (Group 2A), based on limited evidence in humans and sufficient evidence in experimental animals (IARC, 2014). Although carcinogenicity studies concerning tetrachloroethylene have largely focused on the inhalation route of exposure, carcinogenicity has also been reported following oral exposure. The NTP (1986) and JISA (1993) inhalation studies provide the best data for assessment of the risk of tumour development. Linear extrapolation from high-dose studies in rodents to low-dose human exposures might not be representative of the actual risk of exposure to drinking water containing low concentrations of tetrachloroethylene. This is because quantitative differences in metabolism exist when comparing humans with rodents, and saturation of the oxidative pathway in high-dose studies results in a plateau in the generation of oxidative metabolites while driving the metabolism to the GST pathway. For this reason, external dose is not considered to be the appropriate metric for use in risk assessments based on high-dose exposures, and internal dose metrics should be used instead. As described in Sections 8.5 and 9.5, a PBPK model was developed for use in both the cancer and non-cancer risk assessments for tetrachloroethylene. The species-relevant models were applied to studies that were deemed to be relevant for the dose-response assessment to allow for the estimation of various internal dose metrics at each of the exposure levels reported in the studies. Summaries of the approaches for the most relevant health-based values (HBVs) for cancer and non-cancer effects are presented in Sections 10.1 and 10.2, respectively.

10.1 Cancer risk assessment

Detailed results for the cancer dose-response modelling for tetrachloroethylene exposure are presented in Table 3 in Section 9.5.1. As described in the aforementioned section, hepatocellular tumours are the most relevant endpoint for the cancer risk assessment. The default approach for cancer risk assessment for mutagenic compounds (or for compounds that cannot be excluded as being mutagenic) is to perform linear extrapolation to estimate exposure levels that would result in excess risk levels in the range of 10^-5 to 10^-6 (1 in 100 000 to 1 in 1 000 000). The weight of evidence for tetrachloroethylene more strongly supports a non-genotoxic mechanism for hepatocellular tumours; therefore, the TDI approach is considered an appropriate replacement for the default approach for calculation of the HBV for tetrachloroethylene in drinking water. The mouse PBPK model described in Section 8.5 was applied to estimate internal dose metrics in mice from the NTP (1986) and JISA (1993) studies. Benchmark dose values representing a 10% increase in adverse effect over background rates (BMD₁₀) and their lower 95% confidence limits (BMDL₁₀) were then calculated for the internal dose metrics using the U.S. EPA Benchmark Dose Software (BMDS Version 2.2 R67) (U.S. EPA, 2011a). Hepatic oxidative metabolism rate was selected as the most relevant dose metric for hepatocellular tumours, because the generation of oxidative metabolites in the liver is assumed to be most closely associated with the tumours. After obtaining the internal dose BMDs, the human PBPK model was applied to estimate the oral exposures that would be associated with the internal dose levels. The external BMD₁₀ of 6.2 mg/kg bw per day and BMDL₁₀ of 1.7 mg/kg bw per day were obtained from the NTP (1986) study, using a multistage model. To calculate the TDI, the BMDL₁₀ is divided by uncertainty factors to account for inter- and intraspecies variability; only the pharmacodynamic component of the interspecies uncertainty factor (2.5) is used because pharmacokinetic differences between mice and humans were already quantitatively accounted for with the application of the PBPK model. The default uncertainty factor of 10 for intraspecies variability was applied. Finally, an uncertainty factor of 10 was also applied to reflect the fact that the assessment was being performed for a threshold carcinogen, as described in Ritter et al. (2007). The extra level of protection provided by this additional uncertainty factor is also supported by the fact that the specific mode of action of tetrachloroethylene-induced tumours has not been confirmed. Using the BMDL₁₀ value, a TDI can be calculated as follows:

$\text{TDI}=\genfrac{}{}{0.1ex}{}{1.7\text{ mg/kg bw per day}}{250}=0.0068\text{ mg/kg bw per day}$

where:

• 1.7 mg/kg bw per day is the external dose associated with the BMDL₁₀ (using a 95% lower confidence limit) from NTP (1986), as presented in Table 3; and
• 250 is the uncertainty factor (×10 for intraspecies variability; ×2.5 for the pharmacodynamic portion of the interspecies uncertainty factor; and ×10 for the severity of the effect of carcinogenicity)

Using this TDI, the HBV for drinking water can be calculated as follows:

$\text{HBV}=\genfrac{}{}{0.1ex}{}{0.0068\text{ mg/kg bw per day}\times 70\text{ kg}\times 0.2}{6.2\text{ L-eq/day}}=0.0154\text{ mg/L}\approx 0.015\text{ mg/L (15 µg/L)}$

where:

• 0.0068 mg/kg bw per day is the TDI derived above;
• 70 kg is the average body weight of an adult;
• 0.2 is the default allocation factor for drinking water, used as a "floor value", since drinking water is not a major source of exposure and there is evidence of widespread presence in at least one of the other media (air, food, soil, or consumer products) (Krishnan and Carrier, 2013); and
• 6.2 L-eq/day is the daily volume of water consumed by an adult, accounting for multiroute exposure (as described in Section 5.6)

10.2 Non-cancer risk assessment

For effects other than cancer, a TDI can be derived by considering all studies and selecting the critical effect that is most relevant or occurs at the lowest dose, selecting a dose (or POD) at which the critical effect either is not observed or would occur at a relatively low incidence (e.g., 10%) and reducing this dose by an uncertainty factor to reflect the differences between study conditions and conditions of human environmental exposure.

The non-cancer risk assessment is based on neurotoxicity. Neurotoxic endpoints have been demonstrated to be relevant to humans in epidemiological studies performed in environmental and occupational settings and in acute controlled dosing studies; furthermore, the effects are more conservative than those for other non-cancer endpoints (hepatic, renal and reproductive/developmental effects). Neurotoxicity was observed in experimental animal studies, and these were initially considered in the dose-response assessment, but the epidemiological studies are more relevant to humans and demonstrate neurotoxicity at lower levels than the animal studies. The non-cancer risk assessment is thus based on human studies.

As described in Section 9.5.2, a NOAEL of 4.8 ppm for colour confusion was obtained from the key study from the neurological assessment (Cavalleri et al., 1994). Health Canada calculated a BMD₁₀ of 7.2 ppm and a BMDL₁₀ (using a 95% lower confidence limit) of 6.6 ppm using the power model. PBPK modelling was performed using the BMDL₁₀ value to extrapolate from inhalation exposures to equivalent oral doses. Because the parent compound is suspected to be the toxic moiety for neurological effects and peak exposure levels for solvents are often more relevant for neurological effects, the most relevant dose metric was the peak concentrations of tetrachloroethylene. No brain compartment was included in the PBPK model, but partitioning coefficients for tetrachloroethylene in brain are similar to those in kidney (Dallas et al., 1994); therefore, the peak concentrations of tetrachloroethylene in the kidney were used as a proxy for brain values. The external oral dose associated with the BMDL₁₀ was 4.7 mg/kg bw per day.

A 1000-fold uncertainty factor is recommended for the neurological endpoint. The full default uncertainty factor of 10 was applied for intraspecies variability to represent differences in pharmacokinetics and pharmacodynamics within the population, because the PBPK model did not account for variability in either of these factors within the human population. An uncertainty factor of 10 for database deficiency is also applied. Due to limitations in the human neurotoxicity database, the selection of the key study for the endpoint was based on the quality of the study rather than the effect observed at the lowest level. Adverse neurological effects were observed at lower concentrations than those in the Cavalleri et al. (1994) study, but these studies could not be used in the dose-response assessment because they presented exposed participants as a single exposure group or there were differences in characteristics of exposed and control participants that could affect study results. Moreover, the database deficiency uncertainty factor is also applied with the use of an occupational exposure study because the "healthy worker effect" can potentially bias study results by diluting or quantitatively reducing the adverse effects observed.

The third uncertainty factor is a value of 10 to extrapolate from a less than lifetime exposure. Average duration of exposure for participants in the Cavalleri et al. (1994) study was 8.8 years. Although adverse effects were already observed despite a less than lifetime exposure, longer durations could have lowered the exposure levels at which the effects were observed.

The TDI for neurological effects of tetrachloroethylene can be calculated as follows:

$\text{TDI}=\genfrac{}{}{0.1ex}{}{4.7\text{ mg/kg bw per day}}{1000}=0.0047\text{ mg/kg bw per day}$

where:

• 4.7 mg/kg bw per day is the external dose associated with the BMDL₁₀ (using a 95% lower confidence limit) from Cavalleri et al. (1994), as described above; and
• 1000 is the uncertainty factor (×10 for intraspecies variability; ×10 for database deficiency; and ×10 to represent a less than lifetime exposure, as described above).

Using this TDI, the HBV can be calculated as follows:

$\text{HBV}=\genfrac{}{}{0.1ex}{}{0.0047\text{ mg/kg bw per day}\times 70\text{ kg}\times 0.2}{6.2\text{ L-eq/day}}=0.0106\text{ mg/L}\approx 0.010\text{ mg/L (10 µg/L)}$

where:

• 0.0047 mg/kg bw per day is the TDI derived above;
• 70 kg is the average body weight of an adult;
• 0.2 is the default allocation factor for drinking water, used as a "floor value", since drinking water is not a major source of exposure and there is evidence of widespread presence in at least one of the other media (air, food, soil, or consumer products) (Krishnan and Carrier, 2013); and
• 6.2 L-eq/day is the daily volume of water consumed by an adult, accounting for multiroute exposure (as described in Section 5.6).

10.3 Comparison of cancer and non-cancer assessments

The HBV for the non-cancer assessment, which was 0.010 mg/L using human neurological data, is more conservative than the HBV for hepatocellular tumours of 0.015 mg/L. Moreover, there is much greater confidence in the human relevance of neurological effects than that of liver tumours at environmentally relevant exposure levels for tetrachloroethylene. The HBV of 0.010 mg/L that was derived using human neurological data is therefore considered to be sufficiently protective of the carcinogenic effects of tetrachloroethylene.

10.4 International considerations

This section presents the various drinking water guidelines and standards from other international organizations. Variation in these limits can be attributed simply to the year of assessment, or to differing policies and approaches including the choice of key study, as well as the use of different consumption rates, body weights, and allocation factors.

WHO (2003) established a drinking water guideline of 40 µg/L (originally published in 1996) based on a no-observed-adverse-effect level (NOAEL) for hepatotoxic effects from a 6-week gavage study in mice (Buben and O'Flaherty, 1985) and a 90-day drinking water study in rats (Hayes et al., 1986). The Australian drinking water guideline is 50 µg/L, based on the same two studies (NHMRC and NRMMC, 2011).

Tetrachloroethylene is currently regulated in the United States under the National Primary Drinking Water Regulations. The maximum contaminant level (MCL) of 0.005 mg/L is designed to be protective of liver problems and increased risk of cancer, with a maximum contaminant level goal (MCLG) of 0 mg/L (U.S. EPA, 2012e). The U.S. EPA is considering regulating tetrachloroethylene along with up to 15 other VOCs that are known or suspected to cause cancer as a group, under the new Drinking Water Strategy (U.S. EPA, 2011b). This group of VOCs will be the first group of contaminants to be addressed under this new strategy. The U.S. EPA also recently performed an Integrated Risk Information System assessment for tetrachloroethylene. They calculated an oral reference dose of 6 µg/kg bw per day based on the average value for two separate neurological endpoints (Cavalleri et al., 1994; Echeverria et al., 1995), and the concentration in drinking water associated with a 10^-6 excess risk was estimated as 20 µg/L based on hepatocellular adenomas and carcinomas from the JISA (1993) study (U.S. EPA, 2012c).

The Office of Environmental Health Hazard Assessment of the California EPA (OEHHA, 2001) derived a non-regulatory public health goal of 0.06 µg/L for tetrachloroethylene based on hepatocellular carcinomas in mice (NCI, 1977; NTP, 1986) and mononuclear cell leukaemiain rats (NTP, 1986).

• Previous Page
• Table of Contents
• Next Page