| Sponsored by: International
Society of Regulatory |
KEYNOTE PRESENTATION Michael Gough, Ph.D. On behalf of ISRTP and the distinguished scientists who will address us today, I also want to welcome you here. I also want you to know that the opinions expressed in my talk are my own. "It’s Been a Long Time Coming"1 George and Ira Gershwin’s song, "How Long Has This Been Going On?" comes to my mind when I think about dioxin.
I could go on with landmarks in investigations and reviews of the possible health effects of dioxin and chemicals and products contaminated by it. Everyone has been exposed to it. Some workers and residents of Seveso were exposed to ten thousands of times as much dioxin as the average person. The federal government has spent more than a billion dollars on research, and, surely by now, everyone in the United States knows that dioxin is an animal carcinogen and that it is present in the food we eat. Yet, we continue to scarf down 1.1 million servings of Ben & Jerry’s ice cream daily. A Rose is a Rose is a Rose, but is a Dioxin Molecule Always a Dioxin Molecule? Ice Cream: I mention Ben & Jerry’s because that company, along with Greenpeace, publicized the risks of exposure to any amount of dioxin, no matter how small, about a year ago, when the company introduced new packaging with less dioxin in the paper. Certainly Green Peace should know that about 95 percent of human exposure to dioxin comes from the consumption of animal fat and that premium ice creams like Ben & Jerry’s are loaded with animal fat. Confronted with the fact that a single serving of Ben & Jerry’s ice cream contains 200-times as much dioxin as the daily dose that EPA associated with a 10-6 cancer risk, the company responded that their ice cream was safe. To bring the Ben & Jerry’s story in line with current EPA analysis, the risk from Ben & Jerry’s ice cream will have to be increased by about 10-fold. According to the company, its ice cream is safe. I agree with the company. Perhaps nothing illuminates the political nature of dioxin better than Ben & Jerry’s. When the company could burnish its green credentials by damning dioxin, and, by the way, cast doubts on the safety of its competitors’ packaging, it did so. When hoist on its own petard, the company brushed away the risks from dioxin. Mother’s Milk: In my opinion, some of EPA’s behavior about dioxin parallels Ben & Jerry’s. For years, the Agency has warned about the risks from dioxin and it is known about the importance of animal fat in exposures to it. It has breathlessly told the public that human breast milk can have high concentrations of dioxin and that babies may be especially susceptible to dioxin toxicity. Then it turns right around and says that the benefits of breastfeeding far outweigh any risks from it. EPA’s message doesn’t hold together very well. Ben & Jerry’s statements can be interpreted to mean that dioxin from paper is bad, and dioxin in ice cream isn’t. I have another example of variable dioxin toxicity. At EPA’s July meeting for peer review of its re-reassessment, Linda Birnbaum said that a paper, not yet published, from some Dutch scientists showed that background levels of dioxin affected children’s mental development. In the analysis, the children, all of whom have "background" levels of dioxin, were divided into quartiles based on their body burdens. Lo and behold, those in the highest quartile had IQ scores two points (or four, I’m not really sure) lower than the other children. One reviewer commented that those results would lead to predictions of really low IQ scores among populations with really high exposures and that they have not been reported. Alan Smith, who reviewed the paper at Dr. Birnbaum’s request, dismissed it as having no significance. I don’t know what importance EPA attaches to the paper. For whatever reason, Dr. Birnbaum volunteered a comment about the role of breastfeeding in the exposures of children. It turns out that children in the highest quartile of dioxin exposure who had been breast fed didn’t suffer IQ deficits. In my mind, those reported observations lead to one of two conclusions. One, dioxin has nothing to do with children’s development and that parental behaviors, for which breastfeeding is a marker, are far more important. Or, two, the data collected by the Dutch investigators can be interpreted in some sort of politically correct fashion so that dioxin is bad unless it’s dioxin in mother’s milk. Then it’s OK. EPA’s 1994 Dioxin Reassessment We’re here today to listen to a number of experts talk about data, analyses, and conclusions concerning EPA’s re-reassessment of the risks from dioxin. Six years ago, in 1994, EPA released its $2,500 dollars per page, 2,500-page reassessment of dioxin. EPA’s conclusion that dioxin was a significant cancer risk was old news. EPA, in its first dioxin risk assessment in 1984, had made essentially the same estimate of carcinogenic risks. What was new, however, was EPA’s conclusion that environmental concentrations of dioxin were likely to be causing other health effects and diseases ranging from endometriosis and reduced testis size to immune deficiencies and developmental effects. EPA conducted a four-city tour in 1994 to publicize its reassessment. It didn’t mention a letter to the editors of Science in which 17 scientists who had contributed to the reassessment said that the Agency had overreached in drawing conclusions from the science. At the beginning of 1995, EPA delivered the reassessment to the Agency’s Science Advisory Board. In May 1995, the SAB rejected the reassessment. In contrast to EPA, which found evidence for dioxin causing essentially every disease known to man, the SAB made a specific finding. "... the Committee’s conclusion is that chloracne is the only lesion of note clearly established as being related to TCDD exposure."2 Scientific Hypotheses and Picking and Choosing I was a government expert on the SAB committee, and my impression is that EPA staff turned science on its head in writing the reassessment. Karl Popper and others who have thought about the characteristics of science emphasize the importance of designing experiments to test hypotheses. EPA clearly pushes the hypothesis that dioxin is a risk to human health. Equally clearly, neither EPA nor anyone else will or can knowingly expose humans to dioxin to test the hypothesis. But hundreds of reports are available about the effects of dioxin on the health of humans and its effects on animals. Rather than consider the evidence as a whole, which is fundamental to testing any hypothesis, EPA picked and chose the reports that bolstered its case and ignored the rest. Not only did it ignore reports that did not support its conclusions, it ignored biological and biochemical and molecular biology results that indicate the Agency’s preferred straight line extrapolation of effects versus a dose that was incorrect for dioxin. EPA’s 2000 Re-Reassessment William Farland, who had directed the 1994 reassessment, told the SAB that a revision of the reassessment would be available in about six months. He was wrong. Five years passed between the May 1995 SAB meeting and the release of a revised risk assessment in June 2000. The current version is less than a week old, and it is available only on the EPA Web Site.3 For reasons that are unclear to me, EPA is now in a great rush to have its re-reassessment reviewed. EPA says that dioxin emissions decreased 80 percent in eight years between 1987 and 1995 and that human body burdens are less than half of what they were in the last decade. I assume that both are continuing to fall. But EPA says reduced exposures are no reason for complacency because the carcinogenicity of dioxin had been underestimated, and the Agency repeats that dioxin causes many other diseases. "It Means Just What I Choose It To Mean -- Neither More Nor Less"4 I’ve always thought the words "human carcinogen" meant that the substance had been shown to cause human cancer. EPA disagrees. It combines inadequate human evidence with convincing animal evidence and some information about "mechanism of action" to call dioxin a human carcinogen. This is more than rhetoric. It’s twisting words so that they no longer have meaning. It happens all the time. As a lecturer in mathematics at Oxford put it, " 'When I use a word,' Humpty Dumpty said in a rather scornful tone, ' it means just what I choose it to mean -- neither more nor less.' "5 Let me go further down EPA’s "human carcinogen" path. Many animal carcinogens are known mutagens. So far as we know, the mechanisms of mutagenicity are the same in all organisms. Therefore, if there is a smidgen of evidence that a mutagenic chemical is associated with human cancer and evidence that the chemical has caused cancer in animals, by analogy to EPA’s reasoning, that chemical can be called a human carcinogen. EPA’s Remarkable and Too-Little Discussed Finding According to EPA, dioxin is associated with tumors at many sites. That’s in contrast to one of the first things most people learn about cancer. It is not a single disease. Carcinomas, adenomas, sarcomas, lymphomas, leukemias arise from different tissues in different organs. But EPA’s analysis of the human evidence depends on its discovery of a whole new class of chemical carcinogens that cause cancer at all sites. Such a remarkable reported discovery deserves more attention and investigation. History shows that investigating a single proposed chemical-cancer relationship can take many years. Twenty-one years ago, Lennart Hardell and his colleagues reported that as few as five days exposure to dioxin-containing herbicides increased the incidence of soft tissue sarcomas (STS) by five-fold. At the Dioxin 2000 meetings, J.T. Tumisto and his colleagues reported on relationships between body burdens of dioxin and the frequency of STS in 160 STS patients and 100 controls in Finland. In contrast to any prediction that can be made from Hardell’s work, Tumisto found an inverse relationship between dioxin and STS. This study is a test of the hypothesis that dioxin is associated with STS. It shows no relationship; quite the contrary. I don’t know what weight it will be given in EPA’s considerations. Neither do I know how long it would take to investigate, verify, or call into question the idea that dioxin is a carcinogen at everybody's site. Smoking and Other Risks Lung cancers are the major contributor to the total of human cancers associated with dioxin. Everyone knows of the importance of smoking to lung cancer. Information about smoking in the National Institute of Occupational Safety and Health (NIOSH) mortality study of chemical workers is inadequate, based on surveys of smoking in two of twelve factories that were conducted well after compilation of the mortality data had begun. Moreover, in the last few years, EPA has warned about lung cancer as a result of exposures to radon, asbestos, and arsenic, and, don’t forget, environmental tobacco smoke. With all those possibly contributing and unaccounted for exposures, it’s a heroic reach, in my opinion, to conclude that the increased lung cancer in the NIOSH study is dioxin-related. The "International Scientific Community" EPA relies heavily on toxicity equivalence factors (TEFs), which, it says, is accepted by the "the international scientific community." Anyone who looks at the contributors to the TEF scheme may conclude that the "international scientific community" is largely composed of scientists employed by or dependent on regulatory agencies for research support. That’s fine. But that "scientific community" needs a more accurate description. More to Come EPA acknowledges that its interpretation of human studies is shaky and isn’t the only one possible. Almost invariably, EPA goes on to suggest that more or somehow better studies will vindicate its conclusions. How can that be as exposures decline? It can’t. But it allows the EPA to go to the public and say, by golly, we know what’s out there, and we’ll find it. Careful! Slippery Constructions The authors of the re-reassessment rely on conditional verbs and conditional verb constructions. For example: "that there is adequate evidence ... to suspect that humans may respond;" "with potential to produce a broad spectrum of adverse effects;" body burden "closely approaches ... levels at which adverse effects might be expected to occur; "children may be more sensitive;" "evidence supports the health benefits of breastfeeding despite the potential presence of dioxin" (where’d the "potential"come from?); "there is reason to infer;" populations "may be at risk for a number of adverse effects. These may include...." (emphasis added in all quotes). A reasonable rejoinder to each of those statements is, "and it might not." Such conditional statements raise near-impossible barriers for people who question EPA’s conclusions. No matter what specific claim is investigated, others remain, and some claims are so vague that no matter what is done, EPA can respond, "Well, yes, you’re right about that, but it doesn’t address exactly what we had in mind." EPA’s Muted Bottom Line EPA’s Information Sheet 1 at its Web Site concedes, "currently there is no clear indication of increased disease in the general population attributable to dioxin-like compounds."6 Exactly. And very likely there never will be. Equally, because there is no sure evidence that environmental exposures to dioxin have harmed anyone, there can be no evidence that environmental exposures to dioxin have harmed anyone, there can be no evidence that reducing exposures to it will benefit anyone. Government officials have estimated that more than $100 billion has been spent to reduce exposures to dioxin. No matter how much more is spent, there’s no reason to expect any discernible improvement in health. What of the Future? EPA’s re-reassessment, from my reading, glosses over may of the SAB’s objections to the 1994 reassessment. If it’s accepted and put in place, it will raise new problems for the food industry -- including, I suppose, Ben & Jerry’s --and the Food and Drug Administration and the Department of Agriculture. As I see it, the re-reassessment is a slender reed for policy that is based far more on science policy than on science. Footnotes |