Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 95
Executive Summary
Humans live in a complex environment and are often exposed to sequences
or mixtures of toxic materials. The science of dealing with the toxicity of
mixtures is relatively new and is still developing. Most experimental data
relate to the toxic effects of exposures to single materials, yet exposure to
two or more toxic materials might produce greater deleterious effects than
would be anticipated from knowledge of the effects of each of the materials
considered separately. Even as few as two materials can be present in infinite
dose combinations, and relative doses might affect toxicity of the mixture.
Experimental strategies for testing a manageable sample of the infinite com-
binations are obviously needed. Even if the proportions of the constituents
are not of consequence (i.e., if only presence or absence is important), many
combinations can be made from a few constituents. For example, if each
mixture were treated as a separate material, testing all combinations of tox-
icants would be physically and economically impossible. Testing all possible
mixtures of 10 substances (i.e., 2 at a time, 3 at a time, etc.) would require
more than 1,000 tests, even if each substance were tested at only one dose
(present or absent).
An awareness of these problems and the recognition that the biological
mechanisms of toxicity of even single materials are often not known led the
Environmental Protection Agency (EPA) to ask the Board on Environmental
Studies and Toxicology of the National Research Council, through its Safe
Drinking Water Committee, to convene a workshop (1) to review the sci-
entific and experimental issues associated with estimating the toxicity of
mixtures in drinking water, (2) to suggest possible modifications of the current
approaches, and (3) to recommend subjects for future research. This report
95
OCR for page 96
96 DRINKING WATER AND HEALTH
is the product of the workshop. Because the study of the toxicity of mixtures
is relatively new, the Subcommittee on Mixtures, which prepared the report,
confined its considerations to a small number of issues.
APPROACHES FOR PROBLEM RESOLUTION
Several possibilities exist for attacking the problems of testing (and reg-
ulating) mixtures. Research on mechanisms could lead to theoretical modeling
that would exploit data on the toxicity of single materials; that is a desirable
long-term goal. In the absence of knowledge of mechanisms and of accepted
models for combining the toxicities of single materials, some mixtures will
have to be tested. If the numbers of such mixtures can be kept small, testing
might become both operationally and economically possible.
GROUPING CHEMICALS FOR ESTIMATION OF COMBINED RISK
.
A prudent first step in assessing and managing the health risks associated
with exposure to mixtures in drinking water is to reduce the apparent number
of mixtures by developing appropriate methods for grouping the substances
found in water. This report examines two such groups based on similarities
of biologic effects: volatile organic chemicals (VOCs), of which several are
suspected human carcinogens, and organophosphorus compounds and car-
bamates, which inhibit acetylcholinesterase. In addition, this report suggests
some ways to make other logical classifications and thus reduce the testing
needed.
Except for the trihalomethanes (a class of volatile organic chemicals that
are by-products of disinfection), the EPA Office of Drinking Water has
promulgated standards only for single components of drinking water. It is
unlikely that each agent present in drinking water will act in biologic isolation
of every other agent present. Where the mechanisms of toxicity of two or
more toxicants are the same or similar, exposure to several materials, each
at a below-threshold dose (i.e., a dose with a zero response), could amount
to exposure to an above-threshold dose and produce a response.
EPA's general guidelines for evaluating the health risk associated with
exposure to a chemical mixture recommend that, if data are not available on
the complete mixture or a toxicologically similar mixture, a dose-additive
method be used to define a "hazard index" (in effect, an exposure index)
based on measured concentrations and reference doses. The subcommittee
recommends using a modified hazard index that sums similar toxicities,
incorporates multiple toxic manifestations, and suggests, under some cir-
cumstances, the use of an uncertainty factor to allow for possible synergisms.
The subcommittee reviewed earlier schemes for the arbitrary grouping of
substances in its efforts to devise some classification schemes that could
OCR for page 97
Executive Summary 97
facilitate control of mixtures in drinking water. In light of EPA's general
guidelines, the subcommittee suggested the following as possible ways for
EPA to group substances for assessing their combined risk when they are
found in drinking water. Groupings that are more rational for regulatory
purposes might be found.
~ Substances can be grouped according to carcinogenicity. According to
the currently preferred dose-extrapolation models, additivity of response or
risk can usually be assumed for low-dose exposure to a mixture of carcinogens
(at doses with relative risks of less than 1.01~. The subcommittee cautions
that additivity might not apply for carcinogens at high doses, when some
dose extrapolation models are considered, or when one agent is a "pure"
initiator and another a promoter.
· Substances can be grouped according to other toxic end points (such as
specific organ toxicity, peripheral nerve damage, etc.~. It is likely that not
all members of a group will affect an end point via the same mode of action.
· When toxic end points are unknown, substances can be grouped on the
basis of their transformation into similar metabolites with similar reactivity
and stability, although it must be kept in mind that their toxic end points
might differ.
· Substances can be grouped according to structural properties. For toxic
materials that fall into any of these groups, a toxic-equivalence approach
that estimates the combined toxicity of the members of a single chemical
class on the basis of the toxicity of a representative of the class has substantial
appeal. In this approach, one estimates the potencies of the contaminants
belonging to a given class relative to the potency of a representative of the
class. (That is already done for some classes of compounds, such as polycyclic
aromatic hydrocarbons, dibenzo-p-dioxins, and dibenzofurans). Toxic doses
can then be combined according to some weighting procedure using a dose-
additive model.
.
USING PHYSIOLOGICALLY BASED PHARMACOKINETIC
MODELl NG
The new field of physiologically based pharmacokinetics warrants attention
as a first step in applying existing knowledge of mechanisms of toxicity. (In
the realm of toxicity modeling, the term "toxicokinetics" might be more
appropriate.) Unfortunately, little is known about how pharmacokinetic vari-
ables are affected by simultaneous or sequential exposure to multiple chem-
icals. Improved understanding and modeling of the pharmacokinetics of mixtures
should lead to improved methods for estimating the risks associated with
exposure to multiple chemicals in drinking water. Development of appropriate
OCR for page 98
98 OR ~ N K! NG WATER AN D H EALTH
pharmacokinetic models will require considerable theoretical and experi-
mental work.
STATISTICAL APPROACHES FOR REDUCING EXPERIMENTATION
(RESPONSE-SURFACE DESIGNS AND FRACTIONAL FACTORIALS)
Response-surface methods are mathematical-statistical techniques that per-
mit the estimation of the effects of each component of a mixture and the
effects of interactions (i.e., departures from additivity) on the basis of a small
number of experimental points. The 2k factorial approach permits the esti-
mation of possible interactions: each of the k factors (elements in the mixture)
is assumed to be present at two concentrations (one of which may be zero),
and each concentration of each factor is combined with each concentration
of every other factor. Usually, some interactions are of less interest, and that
permits the use of fractional factorial designs, which require still fewer
experimental groups. The availability of modern computer graphics has per-
mitted the plotting of the data, both observed and fitted, in ways that lead
to easier identification of nonadditivity, if there is any.
ASSESSING EXPOSURE
To develop a risk assessment of a drinking water mixture, exposure needs
to be estimated. Drinking water is often not the sole source of exposure to
many of its contaminants. Exposures due to contact with other media and
by other routes must be considered, because they have the potential for raising
total body burden to levels that could be of concern and for providing sub-
stances critical for synergistic interaction with waterborne substances. In-
halation of volatilized drinking water contaminants during cooking, showering,
or other activities is another route by which water can contribute additional
exposure, as is dermal contact in swimming and bathing.
The subcommittee considers it likely that, if a simple analytic process
were developed to provide a summary measure of an entire class of toxi-
cologically similar constituents in drinking water, it would also detect other,
potentially confounding constituents in the water.
APPLICATIONS-PROBLEMS OF EXPOSURE
On a practical level, ambient exposures to mixtures usually involve low
concentrations of the constituents. At concentrations that yield small increases
in relative risks, additive and multiplicative responses are essentially indis-
tinguishable, and additivity is a satisfactory first approximation. For example,
several organophosphorus and carbamate chemicals inhibit acetylcholines
OCR for page 99
Executive Summary 99
terase, and their joint action is assumed to be the sum of their separate actions
on this end point.
The additive approach might need to be modified by incorporating an
uncertainty factor (for possible synergism), which would depend in part on
the information available and the concentrations of the contaminants. If a
great deal of toxicologic information is available on the individual contam-
inants, if toxic interactions are not likely (on the basis of the knowledge
available), or if the concentrations of the contaminants are low, the uncer-
tainty factor might be set at 1 (producing an assumption of simple additivity).
If less is known about the toxicity of individual components and the con-
centrations of the contaminants are higher, the uncertainty factor might be
set at 10. The greater the uncertainty (because of lack of information) and
the higher the concentrations of the contaminants, the larger the uncertainty
factor required to provide an adequate margin of safety. For example, syn-
ergism and antagonism among some anticholinesterases are known to depend
on interference with or competition for metabolic mechanisms of detoxifi-
cation or activation of the anticholinesterases or their precursors. One might
predict that synergism could occur only when the dosages are high enough
for metabolic detoxification to be rate-limiting with respect to toxicity. The
dosages necessary could be lower when other substances inhibit critical path-
ways of detoxification, particularly if the inhibition is of a noncompetitive
type. At the (low) concentrations of anticholinesterase compounds found in
drinking water, it is probably safe to assume no more than additivity of
effect. For such concentrations, the uncertainty factor would be 1.
MIXTURES OF CARCINOGENS
The subcommittee believes that additivity will usually apply to exposure
to carcinogens associated with low risks. However, like the National Research
Council Committee on Methods for the In Vivo Toxicity Testing of Complex
Mixtures, this subcommittee is aware that large exposures to several carcin-
ogens have been shown to produce synergistic interactions-e.g., in a num-
ber of studies cigarette-smoking and exposure to asbestos appear to have
combined to produce a greater than additive risk of cancer although the
mechanisms of carcinogenesis (in this case of asbestos and cigarette smoke-
itself a highly complex mixture) might be different. The subcommittee as-
sumed that exposures to waterborne toxicants, such as VOCs, at low con-
centrations generally are associated with low individual component risks,
although the empirical evidence for such an assumption is sparse. Given that
assumption, it should be possible to estimate the risk associated with exposure
to a mixture of carcinogens by adding the calculated carcinogenic responses
to the individual components of the mixture. The current methods and models
used by EPA to estimate carcinogenic risks in humans on the basis of ex
OCR for page 100
i00 DRINKING WATER AND HEALTH
perimental exposures of animals at high doses are derived from the same
assumption. Additional research should be conducted to provide a firmer
empirical base for the models used.
OTHER ISSUES
The subcommittee is aware of the possibility of substantially increased
risk to some persons associated with even small exposures to chemicals, but
it did not address this extensively. Because of the genetic variability of
humans or because of earlier sensitizing exposures, what is apparently low-
dose exposure for the majority of the population might have serious effects
in a small segment of the same population. Providing advice on how to factor
these complexities into risk assessment was beyond this subcommittee's
charge.
Representative terms from entire chapter:
single materials
