A. The criteria included herein shall be utilized by the Office
of Environmental Health Hazard Assessment Science Advisory Board
Carcinogen Identification Committee (CIC) to identify those chemicals
which are to be recommended for listing as known to the State
to cause cancer. This listing is for purposes of fulfilling the
mandate of the Safe Drinking Water and Toxic Enforcement Act of
1986 ("Proposition 65").
B. These criteria are intended to give the CIC maximal flexibility
in evaluating all pertinent scientific information in determining
whether a chemical is known to the State to cause cancer. They
are intended neither to limit the scope of the Committee's consideration
of all appropriate cumulated scientific information, nor to limit
the use of best scientific judgment available at the time.
C. In evaluating the sufficiency of available data, a "weight-of
evidence" approach shall be used to evaluate the body of
information available for any given chemical. The body of evidence
shall include all evidence bearing on the issue of carcinogenicity
shown through scientifically valid testing according to generally
accepted principles.
D. The Safe Drinking Water and Toxic Enforcement Act of 1986
states that a chemical is known to cause cancer "if in the
opinion of the state's qualified experts it has been clearly shown
through scientifically valid testing according to generally accepted
principles to cause cancer" without further restriction.
Thus if the weight of scientific evidence clearly shows that a
certain chemical causes invasive cancer in humans, or that it
causes invasive cancer in animals (unless the mechanism of action
has been shown not to be relevant to humans), the committee will
normally identify that chemical for listing.
E. The application of causation criteria requires scientific
judgment which can only be based on experience, not only with
the interpretation of epidemiological studies or animal carcinogenicity
experiments in general, but with the circumstances of exposure,
the physical and demographic setting, the nature of classification,
including pertinent clinical and histologic schemata, and the
qualifications of the investigator. Thus, few of the criteria
are amenable to the use of absolute restrictions of either a quantitative
or qualitative nature.
F. Whether evaluating the evidence for carcinogenicity in animals
or humans, CIC members may make judgments utilizing other, more
indirect, scientifically valid observations obtained using generally
accepted methods and principles. Such information may derive from
studies of genetic toxicology or DNA repair using in vitro methods,
cultured mammalian cells, or living prokaryotes, lower eukaryotes,
plants, or insects, although changes induced in whole mammals
must be considered more pertinent. Quantitative variations in
mutagenicity or other short term phenomena cannot be presumed
to always parallel quantitative variations in carcinogenicity,
since not all carcinogens are mutagens. Taken alone, a negative
test can rarely offer strong evidence against carcinogenicity;
although well conducted negative studies can provide important
contributory evidence. Each of the following categories of knowledge
may be pertinent to carcinogen determinations.
Physical and chemical characteristics of the chemical
Absorption, distribution, metabolism, and excretion characteristics
of the chemical
Structure-function and structure-activity relationships
Organ-specific and systemic toxicity, whether after short or long
latency
Protein binding, and cellular receptors
Formation of DNA-adducts by means of chemical binding
DNA repair processes
Effects upon the methylation status of DNA
Mutagenicity of the chemical and its propensity to cause chromosomal
damage
Mutational spectra in observed tumors with known links to environmental
chemicals
A capacity to produce benign tumors known to progress to malignancy
A capacity to produce other effects known to be pre-neoplastic
Epidemiological and experimental studies of such
surrogate outcomes must be held to the same strict criteria as
studies of invasive cancer.
2. Generally accepted principles of scientifically
valid studies of carcinogenesis.
A. Epidemiological studies of carcinogenesis in
humans will be interpreted as showing a causal relationship between
the exposure and the cancer outcome depending on the weight of evidence.
i) Interpretation of the evidence is greatly facilitated
by the availability of the specific details of pertinent studies.
These details would include:
a) The setting and the nature of the population
studied
b) The study design and the sequence of observations
c) The operational definitions of exposure and tumor outcome
d) The means of controlling pertinent bias and confounding
e) The sample size(s) and the details of the analysis, including
statistical testing
ii) The weight of evidence depends upon the degree to which
each of the following propositions can be verified or rejected.
a) The occurrence of the exposure and the occurrence
of the cancer are associated, such that the outcome is shown
to appear more frequently among the exposed than among the unexposed.
b) The observed association cannot be reasonably explained by
chance, based on conventional statistical criteria interpreted
in the context of the number of comparisons made.
c) The observed association is unlikely to be due to any link
between the exposure and other known or presumed determinants
or well-understood predictors of the outcome. The existence
of such other known or presumed determinants does not, by itself,
provide evidence for or against a finding of carcinogenicity.
This criterion can ordinarily not be fulfilled by observations
that link the characteristics of groups rather than those of
individuals.
d) The observed association is unlikely to be explained by biased
working definitions of the exposure or the cancer, or by biased
methods of enumerating either of them.
e) The plausibility of causation is undiminished or is enhanced
by the detailed characteristics of the observed association
as follows; none of these individual characteristics provides
an absolute criterion for or against causality by itself.
1) The strength of any positive association
observed. Credibility is enhanced to the degree that the risk
ratio rises, especially (arbitrarily), above 1.5.
2) The relationship between the dose and/or the duration of
the exposure and the strength of the association. In general,
a direct relationship between these two quantities enhances
the plausibility of a causal explanation.
3) Causality of the observed association is consistent with
what is known of the toxicological and physiologic effects
of the exposure, and with the known causation and pathogenesis
of the cancer in question.
4) The consistency and brevity of the latent period between
exposure and the time of appearance or diagnosis of malignancy.
5) The histological and anatomical description of the tumors
occurring after exposure, including their degree of malignancy
or malignant potential.
6) The biologic credibility of causation as an explanation
for the pattern of time intervals between the period of exposure
and the appearance of the cancer. In general, statistical
variation around a specific period of latency enhances the
plausibility of a causal explanation.
7) The existence of multiple studies, i.e. multiple independent
observations of the same relationship, each of which fulfills
the above criteria. These are especially compelling if studies
differ in respect to study design, population or setting,
measurement technology, analytic strategy, time frame, or
means of estimating what would be expected under the hypothesis
of no association.
8) The absence of any unambiguous observations which are truly
inconsistent with the existence of a causal association. To
be informative, a negative study must be of such quality that,
if positive, it would have added to the weight of evidence.
Such results should be based on definitions of exposure and
cancer outcome which are valid and at least as sensitive and
specific (i.e. have at least as high positive and negative
predictive values) as studies in which an association has
been (or would be) observed. The existence of strong and diverse
indirect evidences such as are listed under General Principle
F above.
B. Studies of carcinogenesis in animals will be
interpreted as showing a causal relationship between the exposure
and the cancer outcome depending on the weight of evidence deriving
from studies employing scientifically valid principles of testing.
i) Interpretation of the evidence from animal
studies is greatly facilitated by the availability of the specific
details of pertinent studies. These details would include:
a) The clear definition and, if a single substance,
the high purity of the agent under test. If pertinent, the means
by which it was collected or extracted, stored, and delivered.
In the case of mixtures, the detailed characterization and composition
of the sample.
b) The route, schedule, and dosage of exposure and the duration
of follow-up. How the dose was monitored, especially in the
case of inhalation experiments.
c) The magnitude of the test dose relative to the maximum tolerated
dose.
d) The species, strain, sex, and age of the experimental animals.
e) The fact and method of animal selection and randomization,
if any.
f) The number of animals in the exposed and in the control groups.
g) The duration of follow-up, the proportion of surviving animals
at risk, and the criteria by which the experiment is terminated.
h) The histological and anatomical description of the tumors
occurring in both exposed and control animals, including the
degree of malignancy or malignant potential of the tumors.
i) The timing of the appearance of tumors.
j) The method of analysis, considering any necessary adjustments
for differential survival, differential examination, historical
as well as concurrent control experience, and the distinction
between progressive tumors and non-progressive tumors found
at autopsy.
ii) The weight of evidence depends upon the degree to which
each of the following propositions can be verified or rejected
with respect to malignancies or tumors of malignant potential.
a) Tumors are found to occur in excess after
exposure to the agent.
b) Tumors appear more frequently in the exposed animals than
in the unexposed comparison group.
c) The observed difference cannot be reasonably explained by
chance, based on conventional statistical criteria interpreted
in the context of the number of comparisons made.
d) The frequency of the unexpected tumors is related to the
dose of the agent.
e) The plausibility of causation is undiminished or is even
enhanced by the detailed characteristics of the observed association
as follows; none of these individual characteristics provides
an absolute criterion of causality by itself.
1) The higher the ratio of tumors in exposed
to tumors in control animals, the more compelling the result,
implying that unusual tumors, occurring in sites rarely affected
under ordinary circumstances, are of special interest.
2) The tumors produced are more aggressive than those occurring
in the absence of exposure. If benign, the tumors are of a
type known to progress to malignancy.
3) Tumors are produced at an especially low dosage of exposure.
4) Tumors occur in unusual variety, or are produced at an
unusually young age or after an especially short interval.
5) Tumors have been found to occur in significant excess (in
order of increasing significance) in the two genders of a
species, in two distinct species, or in two different experiments
carried out in two different laboratories under different
protocols. The following circumstances may constitute exceptions
to this rule:
-- A single study in one species might
be considered to provide sufficient evidence of carcinogenicity,
if the malignant tumors occurred to an unusual degree with
respect to frequency, type, location, age at onset, or low
dosage, or in a strain not otherwise prone to such tumors.
-- Evidence of carcinogenicity in animals deriving from
a single study or from multiple studies incompletely or
inconsistently described might be considered sufficient
if heavily supported by the indirect evidences described
under General Principle F above.
|
Conditions of Use/Privacy
