May 14, 2001
CERTIFICATE AS TO PARTIES. RULINGS AND RELATED CASES

A. Parties and Amici

All parties, intervenors, and amici appearing in this Court are listed in the brief for Appellants, Tozzi, et al.

Disclosure Statement Pursuant to Fed. R. App. P. 26.1 and Circuit Rule 26.1.

Amici Curiae Public Health Scientists are five natural persons who do not issue shares or debt securities to the public.

The following pages identify each scientist who joins in submitting this brief and summarizes his scientific experience in the fields of risk assessment, toxicology, epidemiology, public health and related environmental sciences.

B. Ruling Under Review

References to the ruling at issue appear in brief for Appellants.

C. Related Cases

References to any related cases appear in brief for Appellants.

PUBLIC HEALTH SCIENTISTS WHO ARE SUBMITTING THIS BRIEF

Dr. Christopher J. Borgert, Ph.D.

Dr. Borgert is President of Applied Pharmacology and Toxicology, a firm that specializes in product safety assessment, risk assessment, toxicological study design and causation analysis. Dr. Borgert received a Ph.D. degree in Pharmacology and Therapeutics from the University of Florida College of Medicine. He completed a postdoctoral fellowship in toxicology at the University of Florida Center for Environmental and Human Toxicology, where he served as an external expert reviewer for risk assessments submitted to the Florida Department of Environmental Protection. Dr. Borgert holds a courtesy faculty appointment in the Department of Physiological Sciences, University of Florida College of Veterinary Medicine where he has an active research program in collaboration with other faculty, serves on graduate research committees, and lectures on toxicology and environmental policy. His current research interests include toxicological effects of chemical mixtures and the analysis of drug and chemical interactions. Dr. Borgert served on EPA's federal advisory committee concerning endocrine disruption. He is currently a council member of the International Society of Regulatory Toxicology and Pharmacology.

Dr. C. Jelleff Carr, Ph.D.

Dr. Carr is the founding editor of Regulatory Toxicology and Pharmacology and Professor Emeritus of the School of Medicine, University of Maryland, Baltimore. Dr. Carr has served as a research administrator and consultant with experience in toxicology, pharmacology, medicinal chemistry, and the biomedical sciences, and is currently engaged in studies of existing protocols for the assessment of the carcinogenicity of chemicals in national toxicology programs. Dr. Carr has served as Executive Director of the Food Safety Council (1977-1979) and Director of the Life Sciences Research Office of the Federation of American Societies for Experimental Biology (1967-1977). In Federal service, Dr. Carr served as Chief, Scientific Analysis Branch, Life Sciences Division, in the Office of the Chief of Research and Development of the Department of the Army (1963-1967) and as Chief of the Pharmacology Unit, Psychopharmacology Service Center, National Institution of Mental Health (1957-1963). Dr. Carr has spent 25 years teaching medical students and has published numerous research papers with his colleagues and students. Dr. Carr holds a Ph.D. from the School of Medicine of the University of Maryland, Baltimore.

Dr. Gio Batta Gori, Ph.D.

Dr. Gori is Director of the Health Policy Center, a study group in toxicology, epidemiology, nutrition, risk assessment, and related scientific, policy and regulatory issues. From 1980-88, Dr. Gori served as Director of the Franklin Institute Policy Analysis Center, where he led a program of research in health hazards, risks, and regulatory policies. Dr. Gori formerly served at the National Cancer Institute in various executive positions, including: Associate Scientific Director, Division of Cancer Causes and Prevention (1968-1972); Deputy Director of the Division of Cancer Causes and Prevention (1972-1980); Acting Associate Director of the Carcinogenesis Program (1976-1978). He received the Department of Health, Education and Welfare Superior Service Award in 1976. Dr. Gori is a two-term President of the International Society of Regulatory Toxicology and Pharmacology and has authored over 120 scientific papers. He holds a Doctorate in Biological Sciences from the University of Camerino, Italy.

Dr. John G. Keller, Ph.D.

Dr. Keller is a specialist in General Toxicology and Carcinogenesis. Dr. Keller has served as Member of EPA's Scientific Advisory Board, Subcommittee on Halogenated Solvents. Dr. Keller has reviewed 15 health advisory and criteria documents for EPA's Office of Drinking Water. He also managed preparation of more than 40 Toxicological Profiles for the Agency for Toxic Substances and Disease Registry. Dr. Keller has served as the Director, Carcinogenesis Bioassay Testing for the National Toxicology Program. He was the Founding editor of Drug and Chemical Toxicology, and formerly served as Chair, Gordon Research Conference on Toxicology. Dr. Keller holds a Ph.D. in Animal Physiology, Biochemistry/Pathology from St. Louis University.

Dr. B. Frank Vincent, Jr., Ph.D.

Dr. Vincent currently consults in the areas of toxicology, environmental and human health, and regulatory science. He has expertise in the areas of risk assessment, risk communication and regulatory science. Dr. Vincent is recently retired from Fort James Corp. where his work focused extensively on the assessment and communication of risks associated with dioxin exposures. He was a member of the Industry Risk Assessment Group and the Mechanism of Action Committee which worked to apply advances in cell biology to the risk assessment procedure. Dr. Vincent ha, also lectured in the area of risk assessment. Dr. Vincent received a Ph.D. in organic chemistry from Purdue University in 1962. He is currently the Past President of the International Society of Regulatory Toxicology and Pharmacology.

Page

TABLE OF AUTHORITIES...............................................................vi

GLOSSARY..................................................................................vii

STATEMENT OF INTEREST............................................................2

STATEMENT OF THE CASE...........................................................3

STATUTES AND REGULATIONS.....................................................3

SUMMARY OF THE ARGUMENT....................................................3

ARGUMENT...................................................................................6

BRIEF AMICI CURIAE OF PUBLIC HEALTH SCIENTISTS
IN SUPPORT OF APPELLANTS

This brief Amici Curiae is submitted by five scientists ("The Public Health Scientists") active in risk assessment, toxicology, cancer assessment, public health and related environmental sciences in support of an appeal brought by Plaintiffs in this matter of the District Court's ruling concerning the National Toxicology Program ("NIP") cancer classification for 2,3,7,8tetrachlorodibenzo-p-dioxin ("dioxin").¹ This brief is filed pursuant to Fed. R. App. P. 29 and D.C. Cir. Rule 29, and pursuant to this Court's Order dated January 5, 2001 granting Public Health Scientists' Motion for Leave to Participate Amici Curiae.

Arnici Public Health Scientists submit this brief to emphasize the important scientific and public health issues underlying cancer classifications and to present a scientific perspective concerning NTP's interpretation of its cancer classification criteria for purposes of classifying dioxin. This brief also discusses the appropriate roles of epidemiologic and mechanistic data in classifying potential carcinogens - the substantive scientific issue related to this case.

The scientists who join in this brief are or have been active in research, risk assessment, toxicology, carcinogenicity, epidemiology, public health, and various environmental sciences. In addition, they are leaders in professional societies, advisors to the federal government and scientific consultants to other groups. The Public Health Scientists are dedicated to promoting use of the best available science in regulatory decision-making to ensure health-protective decisions concerning exposures to environmental hazards. The views expressed by the Public Health Scientists represent their independent, long-held positions on matters relevant to the present case.

The present case concerns the recent NTP interpretation of its cancer classification criteria for purposes of defending the agency's decision to classify dioxin a "known" human carcinogen. Because the agency was unable to classify dioxin a "known" human carcinogen based on sufficient epidemiological evidence (which is required by NTP's classification criteria), NTP reinterpreted its criteria to allow it to support a "known" classification with mechanistic information.

As a result of NTP's new interpretation, the line between substances that are "known" to cause human cancers and those "reasonably expected" to cause human cancer has been blurred. This deprives public health professionals of an important tool for distinguishing between potential carcinogenic substances for the purpose of allocating limited cancer prevention resources. For this reason, we believe NTP's actions are imprudent and may harm public health.

In addition, because NTP's new interpretation appears to be an action designed to support a predetermined result, the public may be led to conclude that science can be easily manipulated and ignored.

This fosters distrust of, and cynicism about, regulatory agencies, scientists and science. In sum, amici believe that NTP's actions concerning the dioxin cancer classification may result in harm to public health, to the public trust in government agencies, and to the scientific process, matters that are of great concern to Amici.

Amici adopt the Statement of the Case set forth in the brief of Appellants, Tozzi, et al.

STATUTES AND REGULATIONS

All applicable statutes, etc., are contained in the brief for Appellants.

In 1981, NTP correctly listed dioxin as "reasonably anticipated to be a human carcinogen" based on "sufficient" animal evidence and "limited" evidence from studies in humans. At that time, NTP interpreted "studies in humans" to mean epidemiological studies.² The epidemiological evidence for dioxin is still limited. Nonetheless, NTP has now classified dioxin a "known" human carcinogen. When confronted with the plain language of its classification criteria, which would prohibit a "known" classification for dioxin, the agency first argued that the animal data allowed it to arrive at a "known" classification.³ More recently the agency has taken the position that it is the mechanistic data for dioxin, when combined with limited epidemiological data, that allows the "known" classification. The agency's new interpretation of its criteria is contrary to the plain language of the criteria, renders the criteria nonsensical, and renders the distinction between "known" and "reasonably anticipated" meaningless.

We are concerned that the agency's new interpretation may simply be an attempt to justify a predetermined agency decision to classify dioxin a "known" human carcinogen. Issues related to dioxin have been politically contentious and are the basis of advocacy for many groups which are pressuring agencies, such as NTP, to upgrade dioxin to a "known" human carcinogen. Political expediency, however, should not justify ad hoc alteration of established criteria, and must not take the place of scientifically based decision-making. To ensure reasonably consistent and comparable decisions, agencies promulgate criteria for assessing scientific data. Such criteria are useless if an agency is free to arbitrarily interpret the criteria to satisfy a preconceived decision.

Distinguishing known human carcinogens from reasonably anticipated human carcinogens based on the strength of evidence is sound public health policy. Such distinctions allow the scientific and public health communities, as well as regulators, to prioritize limited resources for the purposes of conducting research and protecting public health. Congress recognized the importance of distinguishing those substances that are "known" to cause cancer in humans from those that are "reasonably anticipated" to cause human cancers. Public Health Service Act of 1978 §301(b)(4) (42 U.S.C. § 241(b)(4)). NTP was charged with making such distinctions and created criteria that would allow it to weigh the scientific evidence and classify substances based on the weight of the evidence. NTP correctly determined when it created its criteria that only sufficient epidemiological evidence would support a "known" classification.⁴ NTP now interprets its criteria to allow it to base a "known" classification on insufficient epidemiological evidence.

NTP's new interpretation is not supported by the plain language of the criteria, which requires the "known" classification to be supported by sufficient epidemiological evidence. Further, from a scientific perspective, "sufficient evidence of carcinogenicity from studies in humans" (which the "known" criteria require) can be obtained only from sufficient epidemiological studies and/or controlled randomized trials i.e., experimental dosing studies in human subjects). But, even if NTP may use insufficient evidence from epidemiology studies along with mechanistic information to achieve a "known" classification, NTP provided no mechanistic information that could be used to draw any conclusion concerning dioxin's carcinogenicity. Rather, NTP identified an early response to dioxin exposure that may or may not be involved in carcinogenicity and is certainly not sufficient to cause cancer.

In sum, the cancer epidemiology data for dioxin are insufficient. According to NTP's criteria this can only support a "reasonably anticipated" classification for dioxin. Even if mechanistic information could somehow make insufficient epidemiological data sufficient, which it cannot, NTP does not have mechanistic data that would allow it to draw any conclusion concerning the potential human carcinogenicity of dioxin.

ARGUMENT

NTP promulgated carcinogen listing criteria for distinguishing between "known" and "reasonably anticipated" human carcinogens. Those criteria state that a "known" classification is appropriate when there is "sufficient evidence of carcinogenicity from studies in humans which indicate a causal relationship between exposure to the agent . . . and human cancer." 61 Fed. Reg. 50,499. NTP has recently interpreted "studies in humans" to include mechanistic information so as to allow the agency to classify a substance a "known" human carcinogen when there is insufficient epidemiological information supplemented with mechanistic information. See NTP's 9^th Report on Carcinogens ("RoC") (April 23, 2001 Supp. to Record ("Supp. R.") at III-58A). We believe NTP's new interpretations cannot be reconciled with the plain language of its criteria, which requires the agency to find sufficient epidemiological evidence before it classifies a substance a "known" human carcinogen.

The meaning of "studies in humans" is clear - it means epidemiological data.⁵ This is the only possible interpretation because it is unethical to conduct mechanistic or other experimental studies in humans using possible carcinogens. Observations in humans may come only after accidental or occupational exposures, and therefore only from observational epidemiologic information. Animal studies and in vitro studies⁶ are not considered studies in humans. NTP recognized this when it limited discussion of such information to its "reasonably anticipated" criteria which do not require a finding of a causal relationship from studies in humans. As discussed below, this common sense distinction between epidemiological and other types of data for purposes of classifying carcinogens is supported by ethical and scientific considerations.

Other language of the "known" criteria strongly supports the conclusion that NTP intended "studies in humans" to mean epidemiological studies. The terms "exposure" and "causal relationship" as used in the "known" criteria are generally associated with epidemiological studies. Indeed, the purpose of epidemiological studies of suspected carcinogenic substances is to attempt to demonstrate a causal relationship between a human exposure and human cancers.⁷ It appears to us that NT? correctly reserved the "known" classification for substances for which sufficient epidemiological information allowed an inference of human carcinogenicity and it used terminology consistent with that intent.

When viewed within the context of the criteria for both the "known" and "reasonably anticipated" classifications, the term "studies in humans" must mean epidemiological studies. To interpret the criteria otherwise would render the distinction between "known" and "reasonably anticipated" meaningless. If, as NTP now claims, "studies in humans" may include mechanistic information, there would be a significant overlap between the "known" and "reasonably anticipated" criteria. For example, according to NTP's new interpretation, a substance for which NTP had limited epidemiology data and relevant mechanistic data could be classified as either "known" or "reasonably anticipated." This interpretation contravenes the purpose of establishing separate categories and contravenes Congress' mandate that NTP distinguish between "known" and "reasonably anticipated" human carcinogens.

NTP has explicitly created a classification for substances for which the agency has less than sufficient evidence of carcinogenicity from epidemiological studies but strong evidence of carcinogenicity from animal or mechanistic studies. NTP classifies a substance as "reasonably anticipated" when (1) human studies do not establish a causal relationship⁸; (2) there is sufficient evidence of carcinogenicity from animal studies; or (3) there is "convincing relevant information that the agent acts through mechanisms indicating it would likely cause cancer in humans." 61 Fed. Reg 50,499. NTP considered the appropriate weight

Glossary Statement of Interest Statement of the Case Statutes and Regulations Summary of the Argument Argument Conclusion

to be placed on animal studies and mechanistic information when it properly chose to explicitly include them within the "reasonably anticipated" criteria rather than the "known" criteria. Interpreting "studies in humans" to somehow implicitly include mechanistic information defies logic and the plain language of the criteria.

NTP correctly established the level of information that would support a "known" classification: sufficient evidence of a causal relationship between exposure and cancer in humans. Phrases such as "sufficient evidence" and "causal relationship" have specific meanings only in science and epidemiology. "Sufficient evidence" suggest a high degree of scientific certainty. Sufficient evidence of a "causal relationship" can only be demonstrated by adequate epidemiological studies.

Scientists and public health professionals consider a variety of scientific information when assessing potential carcinogens. Some of the scientific information is highly predictive, while other information, although useful, is fraught with uncertainty and is of limited predictive value. Because epidemiological studies look for effects directly in humans, uncertainties inherent in extrapolating from animal or in vitro studies to humans are avoided. When properly conducted, epidemiological studies often can provide the most reliable information as to whether there is a causal relationship between an exposure and disease.

Laboratory animal studies can provide useful, albeit less specific and thus less reliable, information concerning the potential effects of a substance in humans. Like epidemiological studies, animal studies examine responses in whole, complex organisms and therefore are far superior to in vitro studies. Still, effects observed in animal studies in which animals are typically dosed over lifetimes and at maximum tolerated doses may not occur in humans at low doses, which creates great uncertainty when one attempts to extrapolate effects observed in animals to humans. Humans may respond differently than laboratory animals because: (1) humans and animals are genetically different and can exhibit very different, species - specific responses to exposures to the same substance,⁹ (2) humans may possess certain mechanisms or biochemical pathways not present in laboratory animals that may prevent expression of a disease; (3) the human diet may modify potential toxic responses; and (4) human exposures to other chemicals may result in induction of protective mechanisms. In addition, the very high doses used in laboratory animal studies may elicit responses that are qualitatively different than those induced in humans exposed to much smaller amounts of substances.

Strong evidence that a substance causes cancer in laboratory animals does raise the possibility that the substance could also be a human carcinogen. For this reason, we believe NIP correctly formulated its criteria to include within its "reasonably anticipated" category those substances for which there is sufficient evidence of carcinogenicity from animal studies. Because of the many differences between humans and laboratory animals, however, such evidence does not rise to the level of demonstrating a causal relationship between exposure and human cancer.

The least useful information for predicting whether a substance causes cancer in humans are in vitro studies, which include mechanistic studies. Such studies may be used to probe narrow questions concerning specific biochemical reactions or mechanisms that may be related to carcinogenicity, but such studies are unable to predict whether a substance will be carcinogenic in humans or in animals. Carcinogenicity is mechanistically complicated and likely involves multiple mechanisms operating within a variety of cells and organs in the body.¹⁰ Such biological complexity is absent in in vitro systems. Other factors that affect the toxicity of substances in humans, such as absorption, distribution, detoxification, elimination, adaptive responses, hormonal influences, immunological responses and numerous other biological activities, are also absent in in vitro experiments. Although in vitro evidence may provide a good screening tool for determining whether further carcinogenicity testing may be warranted, in vitro studies cannot provide evidence of a causal relationship between an exposure to a substance and human cancer.¹¹

In its dioxin classification, NTP appears to claim that somehow a combination of insufficient epidemiological information and equally insufficient mechanistic information can demonstrate a causal association between exposure to dioxin and human cancer. We do not believe that animal and/or mechanistic data can rehabilitate insufficient epidemiological studies for purposes of demonstrating a causal relationship between exposure to a substance and human cancer. In some cases, epidemiological data do not demonstrate a causal association between exposure to a substance and human cancer because the substance is not a human carcinogen. This is often true even when there is evidence of carcinogenicity from animal or mechanistic studies. This is precisely why the "reasonably anticipated" classification was created - to identify substances for which animal, mechanistic or limited epidemiological studies suggest the substance might be carcinogenic to humans.

ATSDR and IARC (the agencies upon which NTP relied when classifying dioxin) have concluded that the epidemiological data are insufficient to demonstrate a causal relationship between dioxin exposures and human cancers.¹² ATSDR stated that "there are limitations in the human database" and that "[b]ecause the human data are incomplete, hazard and risk must be extrapolated across species." ATSDR Profile at 247. IARC concluded that "[t]here is limited evidence in humans for the carcinogenicity of [dioxin]." IARC Monograph at 342.

NTP also found the human data insufficient. In its Draft Background Document for dioxin, NTP stated that "[h]uman studies have found an association between dioxin exposure and cancer . . . ." RBD at RC-1 (emphasis added). It is universally held that an association is not synonymous with a causal relationship.¹³

NTP is more evasive in its recently released dioxin listing for the 9`h Report on Carcinogens Supp. R. at III-58A. Rather than commenting on the sufficiency of the epidemiology data, NTP merely states that it classified dioxin as a "known" human carcinogen "based on sufficient evidence of carcinogenicity from studies in humans, involving a combination of epidemiological and mechanistic information which indicate a causal relationship between exposure to TCDD and human cancer." Supp. R. at III-58A. Still, one must conclude that NTP continues to view the human epidemiology evidence for dioxin as insufficient. Presumably, NTP would not have had to rely on mechanistic information (nor would it have had to reinterpret its guideline to allow such reliance) had it possessed sufficient epidemiologic data for dioxin. Further, NTP continues to rely on the 1997 IARC Monograph and relies on a 1997 paper published by Bertazzi, et al., both of which conclude that the epidemiology data for dioxin is limited.¹⁴

Even if NTP may rely on mechanistic information to demonstrate a causal relationship between dioxin exposures and human cancers, it failed to provide such information. NTP identified in its RoC a single step which may or may not be involved in dioxin carcinogenicity - binding to a receptor (the aryl hydrocarbon ("Ah") receptor). NTP then reasoned: (1) dioxin causes cancer in laboratory animals; (2) those animals have Ah receptors that bind dioxin; (3) humans have Ah receptors that bind dioxin; therefore (4) dioxin causes cancer in humans. This reasoning might work only if it were true that the Ali receptor operated identically in animals and humans, binding to the receptor was necessary to cause cancer in animals and humans, and binding to the receptor was sufficient to cause cancer in animals and humans. NTP's simplistic reasoning must fail for a number of reasons. First, very little is known about how dioxin ultimately causes cancer in animals and even less is known about how it might cause adverse effects in humans. Second, it appears that the Ah receptor may exhibit species-specific differences. Third, it is unknown whether binding the Ah receptor is necessary for cancer to occur, and, in fact, it is known that Ali receptor binding is not sufficient to cause cancer.

Although a few mechanistic responses to dioxin exposures have been identified in animals, it is unknown whether and to what extent any of those responses play a role in animal or human carcinogenicity. Nonetheless, NTP relied on a single response, Ali receptor binding, to construct its support for its dioxin cancer classification. (Ah receptor binding appears to be an early response to dioxin exposure in both animals and humans, and simply means that dioxin attaches preferentially to certain locations in cells.) NTP reliance on Ah receptor binding to predict human carcinogenicity, however, is misplaced because it is not clear what role receptor binding plays in dioxin's carcinogenicity or whether receptor binding is even necessary to cause cancer. It is also unclear whether the Ali receptor is the same in animals and humans. ¹⁵ Further, it has been shown that a "complex series of events regulate the activity of the receptor and it is likely that the differential regulation of these may, at least in part, be responsible for the tissue-and species-specific nature of the response observed in mammals" following dioxin exposure. ATSDR Profile at 233.

Putting aside these important species differences and the fact that the Ah receptor may not be involved in carcinogenicity, Ali receptor binding cannot predict cancer in animals or humans because Ah receptor binding itself does not cause cancer. ATSDR and NTP admit this when they conclude that receptor binding is not sufficient to elicit biological responses. See, e.g., ATSDR Profile at 247; Supp. R. at BI-58B. This is also clear from the fact that humans and animals consume in their diets large amounts of naturally occurring substances (which are not carcinogenic) that bind to the same Ali receptor also preferred by dioxin.¹⁶ Therefore, carcinogenicity is not determined by Ali receptor binding but rather by numerous other biological and biochemical steps that occur after receptor binding, none of which NTP has attempted to describe. Indeed, it cannot describe them because the mechanism for dioxin carcinogenicity is unknown.

NTP's attempt to demonstrate a causal relationship between dioxin exposure and human cancer using mechanistic information sets a dangerous precedent. If the approach NTP used for dioxin was universally applied, most of the substances classified as "reasonably anticipated" because of insufficient epidemiological evidence of carcinogenicity and sufficient evidence of carcinogenicity in laboratory animals would be upgraded to the "known" classification. We believe it would be relatively simple to demonstrate some similar biochemical, molecular or mechanistic response to a chemical exposure that is shared by both animals and humans, especially if one did not have to show that the response was related to carcinogenicity. We are concerned that this would render the distinction between "known" and "reasonably anticipated" meaningless.

Amici are concerned that in the case of its dioxin classification, NTP has altered its criteria to achieve a predetermined dioxin cancer classification that is not supported by the evidence. Such actions harm public health because they prevent a meaningful prioritization of public health risks and, therefore, lead to a misappropriation of scarce public health resources.

It is important to accurately distinguish between substances that are known to cause cancer and those substances for which the evidence of carcinogenicity may be more limited. This allows public health operators to allocate scarce resources in a manner that will do the most good. Public heath officials and regulators might choose to allocate significant resources to limit exposure to "known" carcinogens, whereas they may choose to allocate research funds to further examine substances that are "reasonably anticipated" to cause human cancer. We believe that with its dioxin cancer classification, NTP has rendered the distinction between "known" and "reasonably anticipated" meaningless - an action that would make misappropriation of public health resources inevitable. We are also concerned that NTP's recent misinterpretation of its cancer classification criteria will result in loss of an important public health tool.

After reviewing the evidence for NTP's dioxin cancer classification, Amici can only conclude that the agency conformed its cancer classification criteria to fit a predetermined result. As discussed above, NTP's recent interpretation of its criteria to include mechanistic information as "evidence in humans" is not supported by the plain language of the criteria or by a fair weight of evidence evaluation. In any event, the mechanism for dioxin carcinogenicity is unknown. We believe that NTP's new interpretation of its cancer classification guidelines for purposes of classifying dioxin a "known" human carcinogen may result in harm to public health and the scientific process. Amici respectfully submit that this Court should find that NTP acted arbitrarily and capriciously when it classified dioxin a (mown human carcinogen and should declare that the agency's new interpretation of its cancer listing criteria is unreasonable.

May 14, 2001

WSH\50251.1

CERTIFICATE OF COMPLIANCE

I, Terry F. Quill, hereby certify that the foregoing brief Amici Curiae of Public Health Scientists contains 4,775 words and therefore complies with the type-volume limitation of the D.C. Circuit.

May 14, 2001

CERTIFICATE OF SERVICE

I hereby certify on this 14' day of May, 2001, that a copy of the foregoing brief Amici Curiae of Public Health Scientists in Support of Appellants was mailed by first class mail, postage pre-paid, to the following counsel listed below:

Counsel for Appellants
Charles J. Fromm
Multinational Legal Services, PLLC
11 Dupont Circle
Suite 700
Washington, D.C. 20036

Counsel for Appellees
Stacy M. Ludwig
Assistant U.S. Attorney
Judiciary Center Building
555 4`t' Street, N.W., Room 10-120
Washington, D.C. 20001

Endnotes:

1. The scientists submitting this brief Amici Curiae are: Dr. Christopher Borgert, Dr. Jelleff Carr, Dr. Gio Gori, Dr. John Keller and Dr. Frank Vincent. Biographies for each scientist are set forth in the Rule 26.1 Statement accompanying this brief.

2. NTP modified its criteria in 1996 (61 Fed. Reg. 50,499 (Sept. 26 1996)) but, as fully discussed in Appellant's brief, did not alter the criteria for the "known" cancer classification.

3. See Appellants' brief at 46, note 31; NTP's Draft Background Document for TCDD, September 30, 1997 ("DBD") at RC-1.

4. As discussed more fully in Appellant's brief, prior to its dioxin cancer classification, NTP classified substances "known" human carcinogens only when there was sufficient epidemiological evidence.

5. Epidemiological studies are observations in humans. They may provide convincing evidence of a causal relationship between exposure to a chemical and human cancer. Indeed, NTP has correctly found such convincing causal relationships in numerous epidemiological studies, resulting in a large number of NTP classifications of chemicals as "known" human carcinogens including substances such as asbestos, benzene, radon, tobacco smoking and vinyl chloride monomer, to a name a few.

6. In vitro studies are laboratory studies conducted outside of a living animal and may involve, for example, animal or human cells or tissues. Mechanistic information refers to information related to a discrete biological or biochemical step or steps that may be involved in a biological response, such as cancer. Mechanistic information is generally derived from in vitro experiments.

7. The Reference Manual on Scientific Evidence defines exposure as "The intake into the body of a hazardous material." Federal Judicial Center, Reference Manual on Scientific Evidence, 2d., 2000, at 434. "Intake into the body" suggests the types of exposures seen in epidemiological studies rather that, for example, in vitro studies.

8. According to the criteria, there is limited evidence of carcinogenicity from "studies in humans" when "alternative explanations, such as chance, bias or confounding factors, could not adequately be excluded." The terms "chance," "bias" and "confounding" are typically associated with epidemiological studies. NTP's use of these terms for describing "studies in humans" provides further evidence that NTP intended the term "studies in humans" to refer to epidemiological studies.

9. Animal experiments typically use highly inbred strains of animals that are often particularly sensitive to certain toxic effects. Significant differences are seen even within different species of rodents, different strains of the same species and different sexes of the same strain. For example, female rats were found to be more sensitive than male rats to hepatocarcinogenic effects (i.e., liver cancer) when exposed to dioxin. The difference, however, was not observed in mice. DBD at 7-4.

10. Cancer likely results from a progressive chain of actions and interactions. Each step in the carcinogenic process may be modified by a variety of factors. As a further complication, different carcinogens likely operate through different mechanisms and the same carcinogen might act through different mechanisms in different tissues or in different individuals. Although, some steps in the carcinogenic process have been elucidated for some substances and for some types of cancer, and hypotheses concerning mechanisms for carcinogenicity have been generated, scientists still do not adequately understand the carcinogenic process.

11. A good example is presented by the Ames assay which, using bacterial cells, screens for the possibility that a substance can cause a mutation. This screen is based on the hypothesis that some human cancers may result from mutations. One must be careful, however, not to over interpret the results of such in vitro studies. For example, although it was once thought that the Ames assay could accurately predict whether a substance is a human carcinogen, research has demonstrated that there is a poor correlation between positive results in the assay and cancer.

12. ATSDR is the U.S. Department of Health and Human Service's Agency for Toxic Substances and Disease Registry. ATSDR publishes Toxicological Profiles for Specific Substances and in 1998 published a Toxicological Profile for Chlorinated Dibenzo-p Dioxins (the "ATSDR Profile"). The International Agency for Research on Cancer, or IARC, publishes "IARC Monographs on the Evaluation of Carcinogenic Risks to Humans." In 1997, IARC published a Monograph on Polychlorinated Dibenzo-para-Dioxins and Polychlorinated Dibenzofurans (Vol. 69) (the "IARC Monograph").

13. For example, the definition of "association" in the Reference Manual on Scientific Evidence states that "Association does not necessarily imply a causal relationship." Reference Manual on Scientific Evidence at 387.

14. The abstract to the 1997 Bertazzi paper states that the evidence for dioxin carcinogenicity "is limited" and concluded only that the "association [of observed excess cancers] with dioxin exposure is plausible." Bertazzi, et al., Dioxin Exposure and Cancer Risk: A 15-Year Mortality Study After The "Seveso Accident." Epidemiology 1997; 8(6): 646-652, at 646.

15. There are several different forms of the Ali receptor in mice alone and it is known that the receptor differs between species. ATSDR states that "[t]he extent to which these forms [of Ah receptor] in mice and humans affect the types of responses elicited and the sensitivity to TCDD is unknown." ATSDR Profile at 233.