By Ruthann Rudel and Dick Clapp

Two recent papers by Ruthann Rudel and Julia Brody published in the journal Cancer compiled a list of 216 chemicals shown to cause mammary gland tumors in animal studies and presented a comprehensive state-of-the-science review of environmental factors in breast cancer.  When such important studies are published, it’s typical for the chemical industry or its surrogates to attack them. In this case, Elizabeth Whelan, president of the industry-backed American Council on Science and Health, fired off a response that questioned whether findings from animal cancer studies are relevant to human cancer risk.  Like many who discount current animal cancer studies, though, Whelan didn’t call for something better; instead, she suggested that better peer review  would have kept these papers out of the scientific literature.

Of course, these two papers had already been through a rigorous peer-review process involving scientists knowledgeable about cancer.  Presumably these scientists were aware of something that most scientists understand: We must rely on animal cancer studies because they are the only thing standing between us and a lot more exposure to chemicals that might cause cancer in humans.

Opportunities to identify chemical carcinogens in human studies have been relatively infrequent and are limited to agents that have been studied in the workplace (e.g., bis-chloromethyl ether, vinyl chloride), following medical use (e.g., pharmaceuticals), or after accidental or catastrophic exposures (e.g., ionizing radiation).  Only a small fraction of the thousands of chemicals in use have been directly assessed in human studies, which often have limited power to detect increased cancer risk.  Since occupational studies have not typically included women, these have not been a source of information about breast carcinogens.  Thus, for cancer overall, and for breast cancer in particular, good animal models are our only chance to identify potential human carcinogens and limit exposure.

Current animal cancer tests have important limitations and need to be improved, but it is clear that human carcinogens also cause cancer in laboratory animals (Huff 1993; Huff 1993; Wilbourn et al. 1986).  A typical cancer bioassay conducted by the US National Toxicology Program (NTP) doses male and female mice and rats for 2 years and then counts tumors in all organs.  Usually there are controls and three dose groups with 50 animals per group.  In a study of this size, increased tumor rates of 5-20% can usually be detected with statistical significance.  Smaller increases in tumor rates that are of concern for human population exposures –such as a 0.001% increase or one extra cancer per 100,000 exposed—can only be observed if the number of animals is increased substantially (to several million).  While an individual increased risk of 0.001% seems small, on a population level these risks can be significant.  For example, if there is a small cancer risk associated with a flame retardant in children’s pajamas, and millions of children are wearing these pajamas, this exposure may result in a socially unacceptable number of new cancers. 

Despite these uncertainties, the International Agency for Research on Cancer has concluded that “it is biologically plausible that agents for which there is sufficient evidence of carcinogenicity in experimental animals also present a carcinogenic hazard to humans . . .” and “in the absence of additional scientific information, these agents are considered to pose a carcinogenic hazard to humans.”  Because the standard animal bioassay that is currently considered the best way to identify potential human carcinogens costs  over $2 million per test chemical, most chemicals in use have not been evaluated in animal cancer tests.

Although Rudel et al. (2007) did not evaluate the strength of evidence for carcinogenicity for each chemical in their list, 93 of the 132 chemicals that were reviewed by IARC meet the IARC criteria for “sufficient” evidence of carcinogenicity in animals.  This is consistent with their observation that many of the chemicals cause tumors at other sites as well, and cause tumors in both mice and rats. 

The fundamental question of how well the animal mammary tumor model predicts human breast cancer remains an area of active research and discussion (National Toxicology Program 2006).  As compared with studies of dietary factors and breast cancer, very few breast cancer studies have assessed chemical exposures, so there are limited data to evaluate the predictive power of the animal model.  However, ionizing radiation and hormones such as those used in hormone replacement therapy have been clearly shown to cause breast cancer in humans and mammary gland tumors in animals; and human studies also suggest associations between breast cancer and exposure to PAHs and organic solvents, which also mammary gland carcinogens in animals (Brody et al. 2007). 

Unfortunately, some physicians and epidemiologists involved in cancer prevention have used imprecise statements to describe the potential role of chemical pollutants in breast cancer—confusing a limited database and negative studies for one or two chemicals that have been studied with evidence of a limited role for chemical exposures overall.  Because breast cancer is so common and the environmental chemical exposures hypothesized to affect risk are so widespread, the public health impact of reducing exposures would be profound even if the true relative risks are modest.

Finally, as we await improved screening tests for identifying carcinogenic chemicals, we need to be clear about the implications of ignoring the available data on animal carcinogens.  Before we simply disregard animal cancer tests, we need to ask whether existing surveillance of human chemical exposures and cancer are adequate to protect public health.  Considering the limited exposure monitoring and disease surveillance and the relatively crude tools of environmental epidemiology, we conclude that we can identify only the largest effects, and that most cause-effect relationships would go undetected.  In addition, epidemiologic methods are not preventative, since they do not detect exposure-effect relationships until after large scale exposure and disease have occurred.  We view animal bioassays as a tool for prudent avoidance of unnecessary cancer risks.  If industry-supported groups like ACSH have their way, we would throw the baby out with the bathwater and continue to exposure future generations to avoidable risks.  We agree with the apochryphal story by Mary OBrien about a group of women considering walking across the rocks to cross a river:  Elizabeth Whelan is like the risk assessor advising the women that, based on her calculations, the risk of crossing the river is very small.  She gets mad when the women refuse to take her advice and walk across the rocks.  When she asks the women why, they point to a bridge just upstream.

REFERENCES

• Brody JG, Moysich K, Humblet O, Attfield K, Beehler G, Rudel RA. 2007. Environmental Pollutants and Breast Cancer: Epidemiologic Studies. Cancer, 109 (S11): 2429-2473. [online May 14, 2007, print June 15, 2007]
• Huff J. 1993. Absence of morphologic correlation between chemical toxicity and chemical carcinogenesis. Environ Health Perspect 101 Suppl 5:45-53.
• Huff J. 1993. Chemicals and cancer in humans: First evidence in experimental animals. Environmental Health Perspectives 100:201-210.
• International Agency for Research on Cancer. 2006. Preamble to the IARC Monographs on the Evaluation of Carcinogenic Risks to Humans:World Health Organization.
• Rall DP. 2000. Laboratory Animal Tests and Human Cancer. Drug Metabolism Reviews 32(2):119-128.
• Rudel RA, Attfield KR, Schifano JN, Brody JG. 2007. Chemicals causing mammary gland tumors in animals signal new directions for epidemiology, chemicals testing, and risk assessment for breast cancer prevention. Cancer, 109 (S11): 2397-2428. [online May 14, 2007, print June 15, 2007].
• Whelan EM. 2007. From peer review to fear review. TCS Daily, ed. Nick Schultz. Available: http://www.tcsdaily.com/Article.aspx?id=052207C [accessed June 3 2007]

Ruthann Rudel is the senior scientist in environmental toxicology at the Silent Spring Institute.  She is a leader in exposure, toxicology, and risk assessment related to mammary gland carcinogenesis and endocrine active chemicals.

Dick Clapp is a Professor in the Department of Environmental Health at Boston University School of Public Health, and co-Chair of Greater Boston Physicians for Social Responsibility. Dr. Clapp served as Director of the Massachusetts Cancer Registry from 1980-1989 and worked in two environmental health consulting groups in addition to his teaching and research activities. He was a consultant to the U.S. EPA Science Advisory Board in its 1995 and 2000 reviews of the dioxin reassessment.