| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709, USA
3Correspondence: National Institute of Environmental Health Sciences, 111 Alexander Dr., Research Triangle Park, NC 27709, USA. E-mail: huff1{at}niehs.nih.gov
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
Key Words: environmental carcinogens • assessment of risks • cancer prevention • carcinogenesis bioassay
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
, 2)
. Even though much is known about certain
causes of cancer (e.g., occupational exposures, tobacco smoking, and
treatment with cancer chemotherapeutic agents), the etiologies of most
cancers remain unknown (3
, 4)
. Other widely accepted
factors influencing cancer causation include exposures to occupational,
industrial, and environmental agents (which include synthetic and
natural chemicals), genetic risks, health status, age, diet,
and lifestyle. Several investigators (3
, 5
6
7)
have
concluded that a considerable proportion of cancers are preventable.
Implementation of effective prevention plans (e.g., avoiding hazardous
occupational exposures, reducing tobacco smoking, reducing exposures to
sunlight) could go far to ease this societal cancer burden
(8
9
10)
. Likewise, reducing or eliminating exposures to
cancer-causing agents could have a substantial effect on reducing
cancer incidence, morbidity, and mortality (11)
. A better
understanding of mechanisms of carcinogenesis and of differences in
susceptibility can strengthen the scientific basis for risk reduction
with lowered exposures, especially for sensitive subpopulations.
However, a complete knowledge of the mechanisms contributing to these
diseases is not required to recognize the need to eliminate or reduce
exposures to known cancer-causing substances.
Whereas exposure to tobacco smoke, the most widespread cause of human
cancer, is decreasing in males (at least in certain industrialized
countries), exposure to other environmental carcinogens is increasing
(11)
, and contamination of air, soil, and water is
reaching levels of serious concern (12)
. Unfortunately,
cancer prevention initiatives, despite being the most relevant approach
to reduce cancer incidence, are not considered high priorities in many
health research programs (9
, 13)
.
After the considerable effort devoted to cancer research and
prevention, cancer incidences and mortality from certain cancers
fortunately appear to be decreasing (14)
. For all sites
combined, cancer incidence rates decreased on average 0.7% per year
during the years 1990–1995, and cancer deaths rates for all sites
combined decreased on average 0.5% per year during the same period,
after significantly increasing 0.4% per year during 1973–1990. This
historic first time shift in cancer incidence and mortality is most
welcome, but steady increases in certain cancer rates at several
specific sites continue to be an important public health concern.
In 1976, Ames and colleagues stressed that "we are living in a
sea of chemicals that have not been tested for mutagenicity and
carcinogenicity" and that "thousands of chemicals to which humans
have been exposed have been introduced into the environment without
adequate toxicological testing" (15)
. They were most
concerned with the "halogenated and industrial chemicals to which
humans have unwittingly been exposed" (16)
and that "a
steep increase in human cancer may be the outcome if too many of the
thousands of new chemicals to which humans have been exposed turn out
to be powerful mutagens and carcinogens" (17)
. In 1987,
Ames and colleagues concluded that in addition to epidemiological
findings, "animal bioassays and in vitro studies are also
providing clues as to which carcinogens and mutagens might be
contributing to human cancer" (18)
.
More recently, however, Ames and colleagues have adopted and advocated
the view that exposure to industrial and synthetic chemicals represents
little carcinogenic risk to humans (19
, 20)
. This
perspective has led to their claim of nine ‘misconceptions’ relating
to environmental pollution, health risks from pesticide residues,
methods of identifying chemical carcinogens, and cancer prevention
strategies. In this commentary, we examine several underlying
assumptions in three critical issues that form the primary basis of
these alleged misconceptions: the HERP (Human Exposure/Rodent Potency)
index, the value of animal cancer data, and health risks from natural
chemicals vs. synthetic chemicals. This is followed by a summation of
the facts surrounding each of the nine alleged misconceptions, together
with our perspective on each.
|
THE HERP INDEX: IS THIS A VALID PROCEDURE FOR COMPARING CANCER RISKS? |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
22
23)
.
Although the authors caution that "it would be a mistake to use our
HERP index as a direct estimate of human hazard" (18)
,
they nevertheless use it as a "reasonable strategy" to compare and
rank possible carcinogenic hazards to humans (24)
. Using
this approach, agents with higher values are considered more dangerous
from a public health point of view than are agents with lower values,
and equivalent HERP indexes imply equivalent carcinogenic risks to
humans. For example, Ames states that "Water from the most polluted
well [in Silicon Valley]... has a HERP value orders of magnitude
less than for the carcinogens in an equal amount of soda, beer, or
wine" or in "the average peanut butter sandwich"
(19)
. They further conclude that the carcinogenic
potential of the reference dose of TCDD (6 fg per kilogram per day) is
"equivalent to the alcohol in ingesting one beer every 345 years"
(25)
. Such frivolous comparisons have led Ames to conclude
that "pollution is pretty much irrelevant to cancer"
(26)
. We disagree with this assertion and the methods used
to draw these conclusions.
In the absence of adequate human data, a method that combines cancer potency based on animal carcinogenesis data with information on the prevalence and levels of human exposures in an attempt to estimate cancer risk could be quite useful. Further, in the absence of contrary mechanistic information, it is reasonable to assume a linear dose-response relationship. Indeed, this is what the Environmental Protection Agency (EPA) does in estimating environmental cancer risks. The difference between Ames’ HERP index and EPA’s risk assessment is that EPA takes a more rigorous approach in assessing human exposure and in estimating cancer potency from animal data. Thus, Ames’ approach to using laboratory animal data to evaluate human cancer risk is in sharp contrast to the criteria and scientific judgments used by the National Toxicology Program (NTP), International Agency for Research on Cancer (IARC), and EPA. We can only surmise that the objective is to discredit the use of animal data for assessing human risk.
In addition, there are several noteworthy inconsistencies in
Ames’ use of the HERP index to evaluate human risk. The first involves
the use of cancer risk estimates from ‘high-dose’ animal studies. On
the one hand, Ames states that using laboratory animal studies to
predict cancer risk in humans is a ‘bankrupt’ approach that should be
abandoned (27)
. On the other hand, the TD50 used as the
denominator of the HERP index, an index Ames used repeatedly to compare
relative human cancer risk, is based entirely on data from animal
cancer studies that he criticizes as being unreliable for evaluating
human cancer risks. Thus, it appears paradoxical that Ames criticizes
the use of animal data for public health decisions when he relies on
animal data to generate the HERP indexes for evaluating human risk.
What exactly is the TD50? The TD50 is defined as the daily dose
in mg/kg body wt that, if administered chronically for the life span of
the species, halves the percent of tumor-free animals at the end of a
normal life span (21)
. The TD50 can also be described as
the daily dose that will induce tumors in half of the test animals that
would have remained tumor-free at zero dose. The TD50 is frequently
near the range of the highest dose used in the cancer bioassay or, in
many cases, is even higher than the highest dose used in the animal
cancer study. In the latter instances, the TD50 is obtained by
extrapolating data from ‘high-dose animal cancer tests’ to even
higher doses. Thus, the ‘high doses’ used in animal tests are a
critical part of their HERP index.
Another inconsistency relates to the method of extrapolating risks to
low doses. Ames criticizes linear extrapolation, stating that "the
possible hazard may be markedly less than linearly proportional"
(18)
and that "linear extrapolation (from high doses to
low) just isn’t credible" (28)
. However, relative
rankings of carcinogenic hazards using the HERP index are only relevant
in assessing human health risk to the extent that they also reflect the
relative rankings of these agents at environmentally relevant doses.
And this requires linear extrapolation, which is an inherent assumption
whenever cancer potency comparisons are based on the HERP index.
A related flaw in the HERP index is the use of TD50 values, as opposed to TD10 or TD01 values, for comparing human cancer risk. Using the TD50 ignores all information relevant to shapes of dose-response curves that could be obtained from the available data. For example, two chemicals may have the same TD50 value, but experimental data may show that one has a supralinear dose response whereas the other has a sublinear dose response. In this example, the supralinear chemical would be expected to have a greater effect than would the sublinear chemical at lower doses. Interpretations of the relative effects of these two chemicals based on their TD50 values would incorrectly portray them as having identical low-dose cancer potencies.
Problems associated with the HERP index are not limited to the use of
the TD50 in the denominator. Human exposure doses (the numerator of the
HERP index) may also be difficult or impossible to specify with any
meaningful accuracy. HERP index values have been reported for the same
chemical that vary by as much as 2000-fold with different estimates of
daily human exposure and/or associated risk (15
, 19)
.
Another concern is that daily human exposures are based on a human body
weight of 70 kg (154 pounds), not a realistic estimate for most
Americans. For individuals, especially children, who may consume the
same amount or more of certain foods than do adults (e.g., fruit juices
and peanut butter), the daily exposures would be substantially higher
than that used to calculate their HERP index values. In addition, this
approach does not address differences in susceptibility among
individuals or differences in sensitivity during developmental phases
of life, e.g., the developing fetus. Thus, risks in women or children
may not be adequately captured in evaluations based on exposure
estimates for 70 kg men (29)
.
Another problem with the use of the HERP index is that it regards
single site, single sex species carcinogens as essentially equivalent
to those showing multisite, multispecies effects. For example, to
report HERP index values for d-limonene (a single-site, single species
tumorigen), based on human exposure to orange juice, black pepper,
nutmeg, and mango (24)
, for comparisons with the hazards
of pesticides (e.g., ethylene dibromide, 1,2-dibromo-3-chloropropane)
that are carcinogenic in more than one species is misleading with
respect to scientific judgment of human risk. d-Limonene, shown to
produce kidney tumors only in male rats and frequently cited as a
rodent carcinogen that poses a low likelihood of human cancer risk, has
been unduly amplified to discredit in general the value of experimental
animal data in predicting possible human risk. To the contrary, IARC
places d-limonene in group 3 (30)
, which is among those
chemicals "not classifiable as to its carcinogenicity to humans"
because animal data show a single site, single species effect and there
are no or inadequate cancer data in humans. NTP likewise does not list
d-limonene in its Reports on Carcinogens (31)
. Thus,
neither the IARC nor the NTP concluded that the experimental results
with d-limonene were indicative of a carcinogenic risk for humans. In
addition to d-limonene, numerous chemicals included among the rodent
carcinogens in the CPDB, such as allyl isothiocyanate, benzyl acetate,
crotonaldehyde, petasitenine, senkirkine, and many others, produced
results considered insufficient by IARC and the NTP for assessing human
risk (31
32
33)
.
While the motivation for the establishment of the HERP index appears reasonable, the human exposure assessments and the animal potency values used to calculate HERP numbers are not reliable for comparing potential cancer hazards to humans, especially in potentially sensitive populations such as children. To use HERP numbers to establish regulatory exposure standards would be scientifically unsound.
|
THE VALUE OF ANIMAL CANCER DATA: TESTING CHEMICALS DOES NOT MAKE THEM CARCINOGENS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
. There is little doubt that chemicals cause cancer,
yet fortunately very few appear to do so. (34)
.
The role of long-term animal tests for identifying carcinogens is
primarily to provide reliable information on the capacity of an agent
to produce cancer in mammalian systems and to contribute to decisions
that would lead to better protection of human health. There is also
good evidence that experimental results correlate well with
observations in humans. Animal data on the carcinogenicity of a variety
of chemicals have preceded as well as predicted later epidemiological
observations in humans (35
36
37)
.
The experimental approach to the identification of the causes and
estimations of human cancer risks is improving by making greater use of
our increased understanding of the toxicokinetics of environmental
carcinogens and of new insights on mechanisms of carcinogenesis
(38)
. Strong evidence exists that experimental
results can be extrapolated qualitatively to the human situation.
Nonetheless, expanded efforts are continuing to reduce or eliminate
uncertainties in quantitative risk assessments.
Using data in the CPDB has led to the opinion that 
50% of all
chemicals tested for carcinogenicity in rats and mice are positive in
at least one experiment (18
, 23)
. Unfortunately, the
criteria on which the evidence of carcinogenicity for chemicals in the
CPDB is based are not evaluated critically. A chemical is simply
classified as a carcinogen "if it has been evaluated as positive by
the author of at least one experiment" (20)
. This is at
variance with criteria adopted by the IARC and by the NTP whereby
expert scientific panels carefully and critically assess all available
experimental data before drawing conclusions on the strength of
evidence of potential carcinogenic risk to humans. A large proportion
of chemicals selected for carcinogenicity studies was based on an
a priori suspicion of carcinogenic activity (34
, 39)
.
The high proportion of ‘carcinogenic’ chemicals in the CPDB is in
sharp contrast with the lower proportion of chemicals found to have
evidence sufficient to be classified by IARC as human carcinogens or as
probable human carcinogens. In contrast to the CPDB, the
database of the IARC Monographs indicates that of the 850 chemicals
evaluated within the program, 35% were found to have sufficient
experimental evidence of carcinogenicity in laboratory animals and are
therefore described as carcinogens, 30% exhibited limited evidence of
carcinogenicity, and 35% had inadequate evidence of carcinogenicity
(40)
. Significantly, only 8.6% of these 850 agents are
considered by IARC as ‘carcinogenic to humans’ and 6.7% as
‘probably carcinogenic to humans’. Thus, the CPDB overestimates by at
least three- to fourfold the number of potential human carcinogens.
Ames and Gold (41)
recognize that "carcinogens clearly
are not all the same" and that some agents are positive in two
species while others show carcinogenic effects in only one sex of one
strain, but in compiling tables for ranking possible carcinogenic
hazards from natural and synthetic chemicals, different levels of
evidence of carcinogenicity are treated as if they were equivalent. For
instance, evidence of carcinogenicity of d-limonene and of
1,3-butadiene is treated as if they were actually comparable, when in
fact they are not; d-limonene is ‘unclassifiable’ whereas
1,3-butadiene is a human carcinogen (31)
. Many chemicals
listed in the CPDB are not considered as probably or reasonably
anticipated carcinogens by either IARC or NTP.
|
HEALTH RISKS FROM NATURAL CHEMICALS VS. SYNTHETIC CHEMICALS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
? On the contrary, it is common knowledge that
chemicals existed before the birth of the chemical industry and that
natural chemicals, similar to synthetic chemicals, can be biologically
active or even toxic. Synthetic chemicals differ in their toxicity,
some are very hazardous to human health and others are relatively
inactive. Certainly, the toxicity of synthetic and natural chemicals
may be similar in some cases and different in others. A basic objective
of toxicity testing is to identify the hazardous chemicals in our
environment and characterize their potential to induce adverse effects
in exposed human populations. Carcinogenic synthetic pesticides
constitute a risk no matter what the prevalence of natural chemicals,
which may or may not be true pesticides. There is no interdependence
among natural and synthetic pesticides; consequently, the risks of each
chemical in whichever category one lists them must be evaluated
independently. Obviously, combination and synergistic effects must also
be considered.
Natural carcinogens and anticarcinogens present in food are considered
by Ames and colleagues to be more important determinants of cancer risk
than synthetic carcinogens. Yet no epidemiological data support the
notion that prevention of exposure to natural carcinogens in food
reduces cancer risk, and results from clinical trials have failed so
far to show that specific natural anticarcinogens reduce cancer risk
(43
, 44)
. In contrast, reductions in exposure to synthetic
chemicals in the working environment have been shown to reduce cancer
risks in humans (9)
.
A major component of the estimated daily consumption of ‘natural
pesticides’ is based on the assumed average daily intake of coffee and
of chlorogenic acid, neochlorogenic acid, and caffeic acid contained
therein (20)
. However, large segments of the population do
not consume coffee or plants that have a relatively high content of
caffeic and chlorogenic acid. Most, if not all, so-called natural
pesticides are ingested together with substances that exert a
protective or anticarcinogenic effect, e.g., fruits and vegetables that
are associated with lower risks of cancer. Further, risk is based on
the amounts of carcinogens consumed (dose) as well as the potency of
the putative carcinogenic agent(s). Focusing solely on amounts of
natural vs. synthetic pesticides consumed per day does not address
large differences in consumption or cancer potencies or on coingestion
of other carcinogens and of anticarcinogens.
Attempts to distinguish between naturally occurring and synthetic
chemicals are fraught with difficulties. For instance, it is probably
more appropriate to define as human-made (synthetic), rather than
natural, manufactured items containing ‘natural chemicals’ such as
nickel, chromium, cadmium, or arsenic, and their compounds or tobacco
products, asbestos, and fiberglass, TCDD, gasoline, and many other
substances. This is because human exposures to many of these agents
would not likely occur if it were not for industrial activities.
Tobacco leaves are natural products, but cigarette smoke, as it is
inhaled, may be more accurately described as a manufactured product.
Asbestos is another classic ‘natural’ agent that would not be the
deadly problem it is without human intervention to make it a widely
used and distributed product. Dioxins exist in our environment largely
as a byproduct of waste incineration, paper bleaching, and herbicide
manufacture. Clear distinctions between natural and synthetic chemicals
are not always obvious; hence, the notion that consumption of ‘natural
plant pesticides’ in our diets pose a greater health risk than do
synthetic pesticides residues in our foods is misleading and should be
viewed as an unproved hypothesis. Denying any potential harm of
carcinogenic pesticide residues on fruits and vegetables or in their
products (e.g., juices) does not promote improved public health,
especially for susceptible members of our society (29)
.
|
ALLEGED MISCONCEPTIONS: HAVE THE SCIENTIFIC AND PUBLIC HEALTH COMMUNITIES MISLED THE GENERAL PUBLIC REGARDING ENVIRONMENTAL CANCER RISKS? |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
‘Misconception’ 1: Cancer rates are soaring
Fact: Cancer rates are high
Although cancer rates are not necessarily ‘soaring’, incidence
and mortality rates remain high, and cancer is the second leading cause
of death in the U.S., with more than 2 million new cases per year.
Further, approximately one in every three or four humans in the U.S.
will develop cancer in their lifetime. Thus, it is only prudent to
pursue public health measures that are likely to reduce the risk of
preventable cancers related to exposures to environmental chemicals.
Because a sizable proportion of cancers is preventable, the occurrence
of large numbers of avoidable cancer cases and associated deaths is a
circumstance that society should seek to reduce.
‘Misconception’ 2: Environmental synthetic chemicals are an
important cause of human cancer
Fact: Collective exposures to certain synthetic and natural
chemicals may be important causes of cancer, especially to those
individuals who are at greatest risk
Cancer is thought to be a consequence of genetic and environmental
factors, and there are no accurate estimates of the percentage of
cancers directly attributable to exposure to occupational or
environmental carcinogens. However, even a small percent of 2 million
new cancer cases per year is a significant number. Moreover, the
distinction between natural and synthetic chemicals is not always
obvious or prudent. For example, asbestos, tobacco, and petroleum
products are natural substances that are processed and widely
distributed only because of human intervention. Although people make
their own lifestyle decisions (i.e., voluntary risks resulting from
smoking, obesity, etc.), we expect our air, water, soil, food, and
workplaces to pose little or no risk from cancer-causing pollutants
(involuntary risks).
‘Misconception’ 3: Reducing pesticide residues is an effective
way to prevent diet-related cancer
Fact: Cancer rates would be higher if carcinogenic pesticides
were not regulated
Obviously, a healthy diet should include consumption of fruits and
vegetables. However, pesticides are intended to kill living organisms,
and pesticide residues in or on our food have no known health benefit.
Thus, public health can be better served by eliminating or sharply
reducing pesticide residues, especially those demonstrated to have
carcinogenic potential in laboratory animals or exposed human
populations (e.g., farm workers). Denying any potential harm of
carcinogenic pesticide residues in foods is neither scientifically
defensible nor a prudent public health position.
‘Misconception’ 4: Human exposures to carcinogens and other
potential hazards are primarily to synthetic chemicals
Fact: Humans are exposed to both natural and synthetic mutagens and
carcinogens and reduction of exposure to these chemicals will reduce
cancer risks. The majority of natural chemicals that humans ingest are
not carcinogenic.
The claim that more than 99.99% of ingested pesticides are
natural in origin and less than 0.01% are synthetic pesticides
residues (20)
is both exaggerated and misleading with
respect to human cancer risks. For example, these numbers refer to
pesticides in general and do not distinguish carcinogenic pesticides
(synthetic or natural) from noncarcinogenic pesticides. Second, even if
these estimated values were limited to natural and synthetic pesticides
that are carcinogenic, the comparison is flawed because it ignores
differences in cancer potency, which may be very great among these
chemicals. It should be noted that there is no evidence demonstrating
that all chemicals produced by plants for protection against
environmental insults can be labeled as natural carcinogenic
pesticides. Third, as discussed above, the borderline between natural
and human-made (synthetic) chemicals is also rather blurred. Fourth,
there is no interdependence among natural and synthetic pesticides,
consequently the risk of each chemical must be evaluated independently.
Carcinogenic pesticides constitute a risk no matter what the prevalence
is of natural chemicals.
‘Misconception’ 5: Cancer risks to humans can be assessed by
standard high-dose animal cancer tests
Fact: Both laboratory animal bioassays and human epidemiology
studies are important for identifying human carcinogens
While epidemiological studies provide the most direct
evidence of adverse health effects, design and conduct of these
typically retrospective studies and the analyses of human data are very
costly and time consuming (9
, 11)
, and may often lack
adequate sensitivity. In addition, accurate descriptions of past
exposure circumstances are often difficult to reconstruct. If cancer
risks were to be assessed only from epidemiological evidence of
affected individuals, then the implementation of public health
decisions to reduce or eliminate exposures would require 30 years or
more of prior exposure to provide sufficient numbers of cancer deaths
to ascertain human cancer causation. This practice would have grave
human consequences and is unacceptable.
In laboratory animal studies, exposure levels can be carefully
controlled and potentially confounding factors eliminated. Moreover, in
many instances animal cancer data have preceded or could have predicted
later epidemiological observations in humans (35
, 36)
. All
human carcinogens that have been tested adequately have been shown to
be carcinogenic in animal studies (45)
. Thus, we agree
with the IARC (37)
and the NTP (31)
that in
the absence of comprehensive human data, well-conducted animal studies
are the most definitive means of assessing human cancer risk. In
contrast, Ames disagrees with IARC and NTP, stating that "I don’t
think animal cancer tests are very useful for saying anything about
human cancer" (46)
. However, he does not propose any
alternative experimental approach that could be used to assess human
cancer risk in a predictive manner.
‘Misconception’ 6: Synthetic chemicals pose greater carcinogenic
hazards than natural chemicals
Fact: Natural and synthetic chemicals vary in their
carcinogenic potency
Most occupational and environmental exposures to chemical
pollutants come from synthetic chemical processes and discharges.
Occupationally, synthetic chemicals do pose a greater cancer risk to
workers. The notion that consumption of ‘natural plant pesticides’ in
our diets poses a greater health risk than synthetic pesticide residues
in our foods must be viewed as an unproved hypothesis lacking
scientific credibility. See section on natural and synthetic chemicals
above and response to misconception 4 for more on this issue.
‘Misconception’ 7: The toxicity of synthetic chemicals is
different from that of natural chemicals
Fact: The toxicity of synthetic and natural chemicals
will be different in some cases and similar in others
Although the toxicities of individual chemicals may vary, there
are no basic differences in the basic toxicology of natural and
synthetic chemicals. Characterizing the behavior of a chemical and its
reactivity in biological systems is more important than specifying its
source as natural or synthetic.
‘Misconception’ 8: Pesticides and other synthetic chemicals are
disrupting human hormones
Fact: Hormone disrupting chemicals do affect wildlife and humans,
and the potential risks of environmental endocrine disruptors to humans
need to be more fully determined
The plausibility that environmental endocrine disruptors may
affect humans is based to a large extent on the known carcinogenic
effects of diethylstilbestrol (DES, a synthetic estrogen)
(47)
and other estrogens, as well as observations of
effects in wildlife species exposed to various organochlorine compounds
and endocrine disrupters. Health effects at environmental exposure
levels, especially during critical stages of development, have not been
fully characterized. This is an important and active area of research
that cannot be dismissed by simple denial.
‘Misconception’ 9: Regulating low hypothetical risks advances
public health
Fact: Science-based approaches to risk assessment are necessary for
sound public health decisions
When a chemical shows carcinogenic effects in laboratory animals,
then a prudent public health policy is to limit human exposures.
History has certainly validated this public health strategy. Risks may
be ‘low’ and ‘hypothetical’ only to those who are unaffected or do
not share equally in the benefits of reduced exposure. Concomitant
exposures to low levels of many carcinogens and mutagens must be viewed
with caution, especially since risk evaluations are typically based on
single compounds.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
, 20
, 22
, 24
, 25
, 41
, 48
, 49)
, Ames and colleagues have formulated an
expanding list of alleged myths or misconceptions that discount
1) the relationship between exposure to environmental
hazards and risk of human cancer, 2) the relative importance
of synthetic vs. natural chemicals, 3) the relevance of
using laboratory animals to identify chemical carcinogens, and
4) priorities for cancer prevention and control. However, as
noted above, these declared misconceptions are based on several flawed
assumptions. For example, scientists are not claiming that cancer rates
are soaring, and there are no basic differences in the toxicology
between natural and synthetic chemicals.
More recently, Ames and Gold (50)
have again opined that
regulations aimed at reducing exposure to synthetic carcinogenic
chemicals are actually damaging to public health because they divert
resources and distract the public from the major causes of cancer. We
contend that scientific-based approaches for assessing health risks of
synthetic chemicals are necessary for sound public health decisions,
especially for the most vulnerable subgroups in our society. Disease
prevention strategies cannot rely on agenda-driven proclamations made
by some individuals regarding what constitutes miniscule hypothetical
risks.
Cancer is a collection of multifactorial typically lethal diseases
involving a complex interplay among multiple genetic, environmental,
and host factors. The likelihood of an individual developing cancer in
response to a carcinogenic agent or agents is a consequence of several
interacting factors including the actual exposure circumstances,
genetic factors, diet, lifestyle, health, age, and gender. In addition,
the level of exposure that may be assessed as having a low probability
of risk when considering a single agent may elicit a greater risk when
taking into account multiple and varied exposures to other carcinogens
or modifying agents (4
, 7
, 51)
.
To disregard experimental results as predictive of potential human
hazard may have serious consequences, as past experience has shown. A
dramatic example is that of DES: this synthetic estrogen was given for
years to pregnant women to prevent miscarriages, while ignoring the
experimental evidence of its carcinogenicity; the unfortunate outcome
of this mistake was the high rate of vaginal cancers detected in DES
daughters (32
, 47)
.
The nine alleged misconceptions formulated to support the contention
that exposure to industrial and synthetic chemicals represents only a
negligible cancer risk to humans relies heavily on their HERP ranking
system (19)
. This system is purported to put into
perspective the hazards of selected synthetic chemicals against the
background of natural rodent carcinogens. Our evaluation of the HERP
index revealed several inconsistencies and serious flaws in the
suppositions underlying this ranking procedure, rendering it unreliable
for comparing potential hazards in the general population. Further, the
HERP ranking procedure does not attempt to address differences in
susceptibility among subpopulations (e.g., children, immune
compromised, and genetically susceptible individuals) or the uneven
distribution of exposure in the general population.
A conspicuous proportion of human cancers, varying between 35% and
40%, has been attributed to diet (3
, 4
, 48)
. However, the
evidence of causal associations between specific dietary components and
human cancer are mostly circumstantial, and certainly much weaker than
the causal associations between numerous occupational and environmental
agents and human cancer (9
, 45)
. Humans are exposed
through diet to a great variety of chemicals, both natural and
synthetic, some of which represent a hazard to human health, whereas
others have a protective effect. Pesticides may have an indirect
protective effect by improving the quantity of certain foods, but there
are no data showing that pesticide residues in food provide any health
benefit. Instead, there is evidence that some of them are indeed
hazardous to health.
One considerable difference to note is that exposure to naturally
occurring hazardous chemicals present in our food is generally combined
with exposures to chemicals having a protective effect. Exposure to
industrial chemicals generally occurs without such mitigating effects.
For instance, vitamins and dietary polyphenols may inhibit the
formation of N-nitroso compounds from ingested or endogenously formed
precursors, but would have little effect on exposure to preformed
N-nitroso compounds in the working environment (52)
.
When addressing their ‘major’ risk factors of human cancer
(48)
, Ames and colleagues appear to have ignored the role
of low-dose human exposures to agents that, in many instances, were
identified as carcinogens in the workplace and are therefore generally
labeled as being occupational carcinogens. Occupational carcinogens,
however, such as asbestos, vinyl chloride, benzene, 1,3-butadiene, and
many others do not cease to be carcinogenic when exposures are
encountered outside the workplace environment.
Difficulties in accurately assessing cancer risks due to low-level
exposure to multiple carcinogens do not warrant the denial of these
risks. Levels of ‘acceptable’ risk from exposure to carcinogens are
policy decisions based on public health protection and on a variety of
other considerations, including political, social, technological, and
economic factors. Regulatory decisions for protection against
cancer-causing agents are initially based on predicted increases in
risks of 1 per 1000 for occupational exposures and 1 per million for
environmental exposures. In contrast, Ames suggests that regulatory
efforts should be limited to "worrying about the 1-in-10, 1-in-100,
or 1-in-1000 risks" (53)
. This suggestion translates to
270 thousand to 27 million additional cancer cases in the United
States, an unacceptable level by any standard. The implication that
lesser risks are unimportant is an unacceptable public health position.
Simply claiming that the "very low levels of chemicals to which
humans are exposed through water pollution or synthetic pesticides
residues are likely to pose no or minimal risks" (48)
or
that "the EPA kills people" (54)
(by restricting
permissible pesticide residues in foods, which Ames claims makes fruits
and vegetables economically unaffordable) is counterproductive rhetoric
for the development of sound public health policies.
A reasonable and justified public health approach should emphasize the search for safe chemicals that do not represent any hazard to human health or the environment and attempt to eliminate or sharply reduce exposures to those chemicals that have been identified as hazardous. To screen or test chemicals for their possible mutagenic and carcinogenic activity before or after their introduction into the general environment represents a plain, efficient, and reliably predictive disease prevention strategy. In a similar manner, the systematic screening of chemicals due to become medical drugs, if carefully implemented, decreases the likelihood that new drugs with dangerous side effects will be marketed. The requirement that pesticides be evaluated for their possible adverse effects on human health is not equivalent to declaring that all pesticides are necessarily harmful to humans. It is simply a rational decision guided by the intention of avoiding past mistakes and the view that protection of public health is a goal that society should always pursue.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
Our evaluation of these alleged misconceptions reveals that they are based on flawed and untested assumptions, as well as on faulty scientific judgment. Hence, we believe that in the absence of human data, laboratory animal studies are the most definitive means for assessing carcinogenic potential, and that any reliance on these misconceptions in policy decisions for cancer prevention could have deleterious consequences to human health. To dismiss the role of synthetic or human-made chemicals in the causation of human cancers and to oppose science-based regulatory efforts aimed at reducing exposure to hazardous synthetic and industrial chemicals is a perspective that ignores the value of the precautionary approach for protection of public health. To reject animal carcinogenicity studies for evaluating potential human cancer risks would unfortunately lead to human cancer cases as the only means of demonstrating carcinogenicity of environmental agents. This is unacceptable. We believe that the identification of hazardous chemicals by sound scientific methods and reduction in human and wildlife exposures to such agents will lead to a healthier environment for present and future generations.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Present address: National Cancer Institute, Bethesda, MD 20892, USA.
Received for publication December 16, 1999.
Revision received July 14, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHE HERP INDEX: IS...THE VALUE OF ANIMAL...HEALTH RISKS FROM NATURAL...ALLEGED MISCONCEPTIONS: HAVE THE...DISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. W. Arthur, A. Sanchez-Perez, and D. I. Cook Scoring of predicted GRK2 phosphorylation sites in Nedd4-2 Bioinformatics, September 15, 2006; 22(18): 2192 - 2195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Melnick A Daubert Motion: A Legal Strategy to Exclude Essential Scientific Evidence in Toxic Tort Litigation Am J Public Health, July 1, 2005; 95(S1): S30 - S34. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R.R. Maronpot, G. Flake, and A. J. Huff Relevance of Animal Carcinogenesis Findings to Human Cancer Predictions and Prevention Toxicol Pathol, January 1, 2004; 32(1_suppl): 40 - 48. [Abstract] [PDF]
|
|
|
|
 |
|
 J. Huff Carcinogenicity of Bisphenol A Revisited Toxicol. Sci., December 1, 2002; 70(2): 281 - 283. [Full Text] [PDF]
|
|
|
|
|
|
G. A. Boorman, J. K. Haseman, M. D. Waters, J. F. Hardisty, and R. C. Sills Quality Review Procedures Necessary for Rodent Pathology Databases and Toxicogenomic Studies: The National Toxicology Program Experience Toxicol Pathol, January 1, 2002; 30(1): 88 - 92. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
