Regulation and Risk Assessment of 1,4-Dioxane - Environmental Investigation and Remediation - page 270

model used to develop the current IRIS value. The same parties also argued that a 1,4-dioxane stan-
dard should be established based on the laboratory practical quantitation limit for 1,4-dioxane. *
CDPHE toxicologist Raj Goyal (CDPHE, 2004) rebutted testimony by opponents to CDPHE's
1,4-dioxane regulation. Goyal's arguments included the following points:

•

† on the use of linear and nonlinear
modeling methods support CDPHE's position. Method selection (linear and/or nonlinear)
for low-dose extrapolation should depend upon the availability of information for the oper-
ative mode of action to anticipate the shape of the dose-response relationship. Toxicity
assessments may use both linear and nonlinear approaches if linearity is not plausible and
nonlinearity has support, but a mode of action is not well dei ned. Goyal's brief explains
that an assessment may use both linear and nonlinear approaches if responses appear to be
very different, that is, linear at low dose and nonlinear at high dose, as may occur if there
are separate modes of action at high or low doses.
USEPA's cancer risk guidelines explain that the linear approach assumes that any exposure

USEPA's new draft cancer risk-assessment guidelines

•

dose has some associated risk of causing cancer to develop. Linear dose-response extrapo-
lation implies that risk decreases proportionately with dose below the experimental dose;
linearly extrapolating low doses for cancer risk assessments is generally considered to be
protective of public health (USEPA, 1999).
The nonlinear extrapolation approach assumes that cancer effects are only likely to occur

•

if exposure dose is high enough to overwhelm the protective capacity of the receptor.
Cancer effects may be a secondary effect of toxicity or of an induced physiological change
that is itself a threshold phenomenon. Nonlinear dose-response relationships therefore
imply that responses will decrease much more quickly than linearly with dose (USEPA,
1999). When the mode of action is understood well enough to support a nonlinear extrapo-
lation, a dose-response curve is not estimated below the point of departure. Instead, an
analysis is used to describe the margin of exposure (or distance) from the point of depar-
ture to a dose where there would be a little concern for cancer.
The CDPHE rebuttal brief concludes that the lack of understanding on the mode of action •

of 1,4-dioxane justii es using both linear and nonlinear extrapolation approaches. Therefore,
the USEPA IRIS toxicity value [the cancer slope factor (CSF)] ‡ is not invalidated. The
exact mode of action of 1,4-dioxane cancer and noncancer effects has not been established
because 1,4-dioxane toxicokinetics and metabolism, as well as the toxic potential of its
metabolite(s), are not yet delineated (CDPHE, 2004). Possible metabolites of 1,4-dioxane
include the compounds 1,4-dioxane-2-one, -hydroxy-

ethoxyacetic acid (HEAA), diglycolic acid, diethylene glycol, and oxalic acid. CDPHE
notes that it is not known whether 1,4-dioxane is directly metabolized to HEAA or whether
1,4-dioxane-2-one is the principal metabolite, which then undergoes hydrolysis to HEAA.
The two metabolites, HEAA and 1,4-dioxane-2-one, are readily interconvertible depend-
ing on the pH of the solution; HEAA is produced at alkaline pH.
Genetic toxicity studies indicate that 1,4-dioxane is a weak genotoxin (causing direct

β -hydroxyethoxy acetaldehyde, β

•

damage of genetic material) and a strong promoter of tumors. CDPHE concludes that the
mechanism or mechanisms by which 1,4-dioxane exerts its carcinogenic effects remain

* In 2003, when the Michigan MCL for 1,4-dioxane was under consideration, the example cited for using the PQL as the
basis for setting a standard was Massachusetts, which at that time had a 50 ppb advisory level based on a 50 ppb PQL.
† USEPA (1999, 2005).
‡ The cancer slope factor (CSF) estimates the carcinogenic potency of a chemical, characterized as a plausible upper bound
on the increased human cancer risk from lifetime exposure to an average daily dose of 1 mg/kg. Therefore, the slope
factor is expressed as average dose in units of (mg/kg d) -1 . Lower values of CSFs rel ect lower carcinogenic potency of a
chemical and result in higher drinking water standards.