Federal Contaminated Site Risk Assessment In Canada Part I: Guidence on Human Health Preliminary Quantitive Risk Assessment (PQRA)
Appendix B
Essentially Negligible Cancer Risk for Contaminated Site Risk Assessment
When assessing risks posed by exposure to carcinogenic substances, regulatory agencies such as Health Canada and the United States Environmental Protection Agency (U.S. EPA) assume that any level of exposure (other than zero) is associated with some hypothetical cancer risk. As a result, it is necessary for regulatory agencies to specify a level of carcinogenic risk that is considered acceptable, tolerable, or essentially negligible.
In the 1970s, the U.S. Food and Drug Agency (FDA) was the first agency to address this issue, adopting a risk level of 1-in-1-million (10-6) as the incremental cancer risk for carcinogenic residues in foods that was considered to be "essentially zero" (Kelly, 1991). The origin of this "essentially zero" risk level was purely arbitrary. Since then, the 10-6 risk level has become commonplace in the regulation and management of environmental contaminants, with the strongest endorsement coming from the U.S. EPA, which employs 10-6 as its primary risk benchmark for "acceptable" exposure to carcinogens within the general population.
Although a 1-in-1-million (10-6) cancer risk is the most frequently used risk level for the management of risks posed by environmental (including soil) contamination, many agencies and provinces, including the U.S. EPA, identify a range of increased cancer incidence risks; generally, from 1-in-10,000 (or 1 x 10-4) to 1-in-1,000,000 (or 1 x 10-6) is considered an acceptable risk range depending on the situation and circumstances of exposure (Graham, 1993; Kelly, 1991; Lohner, 1997; Travis, 1987; U.S. Environmental Protection Agency (U.S. EPA), 1991).
In contrast, many industrial standards for workplace environments (such as those of the American Conference of Governmental Industrial Hygienists [ACGIH], 2002) offer a protection to only the 1 x 10-3 level or higher of risk (e.g., a risk of 1 x 10-2, or 1-in-100, is a 1 percent chance). This higher cancer risk is "accepted" in workplace environments because it is often technologically or financially infeasible to reduce exposures to even lower levels, and the nature of exposure is generally deemed to be informed and "voluntary" in the workplace. The U.S. Supreme Court has upheld the industry basis for such standards (Graham, 1993).
In establishing generic Canadian soil quality guidelines, the Canadian Council of Ministers of the Environment (CCME) (1996) prescribed the 10-6 level of risk as being essentially negligible. This was established as the lowest common denominator amongst provincial and federal agencies participating in the CCME guidelines derivation process. However, the CCME (1996) acknowledges that the designation of negligible cancer risk is an issue of policy rather than of science, allowing different agencies to establish such a policy consistent with their respective environmental regulatory agendas. To that end, Health Canada, when publishing human health soil quality guidelines in support of the CCME process, applied the concentration of carcinogenic substances in soil associated with risks ranging from 1-in-10,000 (10-4) to 1-in-10,000,000 (10-7) (see Health Canada, 1995, for example).
Health Canada (formerly Health and Welfare Canada [HWC], 1989), as the federal advisor on environmental health issues, has established that a cancer risk in the range of 1-in-1-100,000 (10-5) to 1-in-1-1,000,000 (10-6) is "essentially negligible" for carcinogenic substances in drinking water. Although published Health Canada advice on this issue has been restricted to exposures via drinking water, the 10-5 risk level has been widely accepted by federal agencies and others involved with contaminated site risk assessment. This level of risk was deemed essentially negligible for risk assessments being conducted in Sydney, Nova Scotia, for soil-borne carcinogenic contaminants associated with the Sydney Tar Ponds, for example (JDAC Environment Ltd., 2002).
The Atlantic Provinces (NS, NB, PEI, and Nfld./Lab.) have implemented a common approach to contaminated site risk assessment known as Atlantic Risk-Based Corrective Action (RBCA) (Atlantic Partnership in RBCA Implementation [Atlantic PIRI], 1999). Within that common risk assessment / risk management framework, an acceptable or essentially negligible cancer risk level of 10-5 has been adopted.
The background incidence of cancer in Canada and the U.S. is high, relative to a 10-5 or 10-6 risk level. The lifetime probability of developing cancer in the U.S. and Canada is approximately 0.4, or 40% (National Cancer Institute of Canada [NCIC], 2001; National Cancer Institute [NCI], 1999). Thus, an excess or incremental cancer risk of 1 x 10-5 increases a person's lifetime cancer risk from 0.40000 to 0.40001.
Some unknown proportion of this "background" cancer incidence is believed to be associated with exposure to environmental pollutants. However, a 10-5 incremental (i.e., over and above background) cancer risk represents only a 0.0025% increase over background cancer incidence; an increase that would be undetectable using available epidemiological data and statistics, particularly in smaller populations that may reside near contaminated sites.
Hypothetical incremental cancer rates associated with carcinogenic substances at contaminated sites are estimated from cancer "slope factors" or "unit risks" derived from human epidemiological studies and animal cancer bioassays. Generally, the incidence of cancer for occupationally exposed adults or laboratory animals (both of which are exposed to dose levels far in excess of exposure levels in the general population or in populations residing near contaminated sites) is plotted against the exposure dose (often standardized for exposure duration, particularly for occupational studies), and a dose-response curve is fitted to those data. This dose-response curve is then extrapolated from the study exposure range down to a dose of zero, with the assumption that there is no threshold below which cancer will not occur. In the U.S. (Crump, 1996), low-dose extrapolation is achieved through application of the linearized multistage model, a statistical model that can describe both linear and non-linear dose-response patterns, and that produces an upper confidence bound on the linear low-dose slope of the dose-response curve. Health Canada often applies this same methodology for the derivation of the TC05 (the concentration in air or water found to induce a 5% increase in the incidence of, or deaths due to, tumours considered to be associated with exposure; see Health Canada [1996]) or the TD05 (the dose found to induce a 5% increase in the incidence of, or deaths due to, tumours considered to be associated with exposure). Health Canada may also apply a model-free low-dose extrapolation method (Krewski et al., 1989), making no a priori judgments regarding the shape of the dose-response curve in the low-dose range. The model-free approach can also provide an upper bound estimate on the slope of the dose-response curve in the low-dose range. These upper bounds on the dose-response curve become the slope factors or unit risks employed for the estimation of hypothetical cancer rates. As such, it is believed (but not proven) that the slope factor or unit risk for carcinogenic substances will overestimate the true cancer incidence associated with low-dose exposure to environmental pollutants, such as from contaminated sites (Kelly, 1991).
Given the conservatism (safety) margin associated with the derivation of cancer slope factors and unit risks, and the negligible impact of a 1-in-100,000 incremental risk level for contaminated site exposures, a cancer risk level of 1-in-100,000 (1 x 10-5) is recommended for the purposes of assessing and managing federal sites contaminated with carcinogenic substances.
References
American Conference of Governmental Industrial Hygienists (ACGIH). 2002. TLVs and BEIs. ACGIH, Cincinnati, OH.
Atlantic Partnership in RBCA Implementation (Atlantic PIRI). 1999. Atlantic RBCA Reference Documentation, Version 1.0. Atlantic PIRI. April 1999.
Canadian Council of Ministers of the Environment (CCME). 1996. A Protocol for the Derivation of Environmental and Human Health Soil Quality Guidelines. Report CCME EPC-101E, CCME. March 1996.
Crump, K.S. 1996. The linearized multistage model and the future of quantitative risk assessment. Hum. Exp. Toxicol. 15(10): 787-798.
Graham, J. 1993. The legacy of one in a million in risk in perspective. Harvard Center for Risk Analysis. Risk in Perspective 1:1-2.
Health Canada. 1995. Canadian Soil Quality Guidelines for Contaminated Sites. Human Health Effects: Inorganic Arsenic. Air and Waste Section, Environmental Health Directorate, Ottawa. Final Report. March 1995.
Health and Welfare Canada (HWC). 1989. "Derivation of Maximum Acceptable Concentrations and Aesthetic Objectives for Chemicals in Drinking Water." In: Guidelines for Canadian Drinking Water Quality - Supporting Documentation. Health and Welfare Canada. Prepared by the Federal-Provincial Subcommittee on Drinking Water of the Federal-Provincial Advisory Committee on Environmental and Occupational Health. Ottawa, Ontario.
JDAC Environment Ltd. 2002. Human Health Risk Assessment - North of Coke Ovens (NOCO) Area, Sydney, NS. Contract report submitted to Public Works and Government Services Canada.
Kelly, K.E. 1991. The Myth of 10-6 as a Definition of "Acceptable Risk". Presented at the 84th Annual Meeting and Exhibition of the Air and Waste Management Association, Vancouver, BC, June 16-21.
Krewski, D., D. Gaylor, and M. Szyszkowicz. 1991. A model-free approach to low-dose extrapolation. Environ. Health Perspect. 90: 279-285.
Lohner, T.W. 1997. Is 10-6 an appropriate de minimus cancer risk goal? Risk Policy Report, April 18, 1997, pp. 31-33.
National Cancer Institute (NCI). 1999. SEER Cancer Statistics Review, 1973-1996. NCI, National Institutes of Health, Bethesda, MD.
National Cancer Institute of Canada (NCIC). 2001. Canadian Cancer Statistics 2001. NCIC, Toronto, Canada. Available online at: http://66.59.133.166/stats/maine.htm
Travis, C.C., et al. 1987. Cancer risk management: a review of 132 federal regulatory agencies. Environmental Science Technology 21: 415-420.
U.S. Environmental Protection Agency (U.S. EPA). 1991. Risk Assessment Guidance for Superfund: Volume 1 Human Health Evaluation Manual (Part B, Development of Risk-based Preliminary Remediation Goals). Publication 9285.7-01B. Office of Emergency and Remedial Response, U.S. EPA, Washington, DC.
|
