Dames & Moore, 1999 - USDA Forest Service

are based on different mechanistic assumptions and expressed in different units. The two approaches are
discussed below. Available toxicity criteria for humans are summarized in Table 7.1-38, and brief
toxicological profiles for the IHSs are presented in Table 7.1-44.
Non-Carcinogenic Effects
Non-carcinogenic cleanup levels were calculated using reference doses (RfDs) developed by USEPA. An
RtD is an estimate of the daily lifetime exposure level to humans (expressed in units of mg of chemicalkg
of body weightjday), including sensitive subgroups, that is likely to be without appreciable risk of
deleterious effects (USEPA, 1989). RfDs are usually derived from oral exposure studies with the most
sensitive species, strain and sex of experimental animal known, the assumption being that humans are as
sensitive as the most sensitive organism tested. They are based on the assumption that thresholds (exposure
levels below which no adverse effect is expected) exist for non-carcinogenic effects, and incorporate
uncertainty factors to account for the required extrapolations from animal studies and to ensure protection of
sensitive human subpopulations. RfDs for constituents considered in the HHRA were obtained, whenever
possible, from CLARC II. Table 7.1-38 summarizes both oral and inhalation RfDs and the toxic effects
endpoint for IHSs.
Carcinogenic Effects
In contrast to non-carcinogenic effects, USEPA typically assumes that there is no threshold for carcinogenic
responses; that is, any dose of a carcinogen is considered to pose some finite risk of cancer. The evidence
for human carcinogenicity of a chemical is derived from two sources: chronic studies with laboratory
animals, and human epidemiological studies where an increased incidence of cancer is associated with
exposure to the chemical. As with the non-cancer toxicity studies, the most sensitive labratory species is
generally used in cancer protocols.
Since risks at the low levels of exposure usually encountered by humans are difficult to quantify directly by
either animal or epidemiological studies, mathematical models are used to extrapolate from high
experimental to low environmental doses. The slope of the extrapolated dose-response curve is used to
calculate the cancer slope factor or potency factor (CPF), which defines the incremental lifetime cancer risk
per unit of carcinogen (in units of risk per mg/kg/day). The linearized multi-stage model for low-dose
extrapolation most often used by USEPA (USEPA, 1986a) is one of the most conservative available. and
leads to a upper-bound estimate of risk (the upper 95 percent confidence limit on the modeled animal dose-
response slope). The probability that the true risk is higher than that estimated is thus only 5 percent.
Actual risk is likely to be lower, and could even be zero (USEPA, 1986a).
Each tested chemical is assigned a weight-of-evidence classification that expresses its potential for human
carcinogenicity. The USEPA's weight-of-evidence classification system is shown below:
G:\~pd#uW)S\rrponrUloldco-2\n174~da 7-25
17693-005-019Uuly 27.1999;5: 16 PMDRMT FINAL RI REPORT