Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Keeping Pace with Science and Engineering. 1993. Pp. 221-242. Washington, DC: National Academy Press. The Dioxin TCDD: A Selective Study of Science and Policy Interaction John A. Moore, Renate D. Kimbrough, and Michael Cough The term dioxin refers to any one of 75 chemicals that have the same basic chemical structure but vary in the number and location of their chlo- rine atoms. This case study focuses only on TCDD, which is shorthand for the prototypical dioxin, 2,3,7,8-tetrachlorodibenzo-p-dioxin (Figure 1~. Re- lated compounds, such as other dioxins or dibenzofurans (a structurally related class of chemicals) did not influence the issues highlighted in this paper and are omitted from most of the discussion. Current scientific thought is that dioxin and dibenzofurans may exert their biological (toxico- logical) effects through a common mechanism. This paper briefly describes three activities that involved TCDD to illustrate the interface between scientific information and policy decisions. These are the registration of 2,4,5T for use as a pesticide; the military use of Agent Orange and the Veterans Administration (VA) activities relevant to that use; and setting state water quality standards for TCDD in the Clean Water Act. Figure 2 summarizes the sequence of events. MEDICAL AND SCIENTIFIC DATA The dioxin TCDD is not a commercial product. It is a contaminant formed in the manufacture of certain commercial chemicals and a product of the combustion of certain materials. Its toxic potential was first reported by Schultz in 1957. He identified TCDD as the causative agent in the development of a skin disease, chloracne, in chemical workers involved in the production of 2,4,5-trichlorophenol. 221 -
222 MOORE, KIMBRO UGH, AND GO UGH Cl Cl 2 3 4 '0` - o 9 6 Cl Cl FIGURE 1 Basic chemical structure of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Trichlorophenol was a feedstock for the synthesis of other chemicals such as hexachlorophene and the herbicide 2,4,5-T (2,4,5 trichlorophenoxyacetic acid). The amount of TCDD formed during the production of 2,4,5 tri- chlorophenol depended on the chemicals used in synthesis and the tempera- ture of the process; some TCDD levels were in the range of 1 to 30 parts per million. The amount of TCDD contained as inpurities in hexachlorophene and 2,4,5-T depends on how much was present in the original trichlorophenol. Because hexachlorophene had medicinal uses, a purified form of trichlorophenol was used in its synthesis and TCDD concentrations were in the range of only a few parts per billion. Animal Studies Concern about the health effects of TCDD remained narrowly focused on relationships between occupational exposures and chloracne until 1970 when teratology studies with 2,4,5-T indicated that exposure to pregnant mice resulted in the production of cleft palates in offspring (Courtney et al., 1970~. Additional studies (Courtney and Moore, 1971; Sparschu et al., 1971) soon revealed that it was the TCDD impurity in 2,4,5-T that ac- counted for the vast majority of the teratogenic or fetotoxic responses in mice and rats. In subsequent years a number of studies confirmed and extended the developmental toxicity observed with TCDD. In addition, studies with TCDD that spanned three generations reported adverse repro- ductive effects (Murray et al., 1979~. Several years later, a number of male veterans who served in Vietnam expressed concern that birth defects in their children were a consequence of their military exposure to Agent Orange, a mixture of the herbicides 2,4,-D (2,4 dichlorophenoxyacetic acid) and 2,4,5-T. To investigate the possibility of a causal association between exposure to Agent Orange and birth defects, an extensive study was performed in male mice. In these studies, high doses of Agent Orange, and TCDD, were employed sufficient to cause Mild toxicity in the mice. Nonetheless, fertility was not impaired and no
THE DIOXIN TCDD Science and Engineering -TCDD identified as causing the occupational skin disease, chloracne 1 945 1 _ 1 950 - Rachel Carson publishes Silent Spring - ~ 1 955 1960 - Herbicides first used in Vietnam _ ~~ _ - ·- - Use of Agent Orange begins in Vietnam 1 965 - ~ ~ r - Spraying of Agent Orange peaks in - TCDD from chemical waste implicated in _ _ u Vietnam 19 70~ - USDA reduces domestic uses of pesticide ~ ,, 1975 - initial laboratory studies report 2,4,5-T as teratogen. Role of TCDD established 223 Policy and Regulation - 2,4,5-T registered as a pesticide by USDA illnesses in horses (and possibly children) in Missouri - Concern about miscarriages in Alsea, Oregon - TCDD a carcinogen in laboratory studies - Love Canal - Alsea, Oregon II study associates miscarriages with herbicide spraying - Epidemiology study associates soft ~ tissue sarcoma in humans with I exposure to herbicides - Seveso 5-year report-no ill effects - Times Beach, Missouri buyout - Ad hoc expert committee advises EPA that - linearized model for cancer is inappropriate - EPA scientific group recommends moderating , cancer risk estimate I - EPA SAB finds no new science data to support change in cancer risk estimate; critical of current cancer model - Banbury Conference supports a receptor- mediated event for dioxin activity - Fingerhut epidemiology study published - EPA initiates review of dioxin risks - P.L. 102-4, Agent Orange Act" enacted - IOM/NAS starts review of evidence associating herbicide exposure with disease in Vietnam veterans 1 1 980 1 ,_ _ ~, - . 19 85 _ . L - Use of Agent Orange ends in Vietnam - EPA announces it was considering cancellation of all uses of pesticide 2,4,5-T - · - EPA suspends use of 2,4,5-T. Vietnam veterans start class suit - Vietnam veterans class action suit settled for $180 million · EPA cancels 2,4,5-T registration ~ _ _ _ - · - Vietnam veterans compensated for soft · 1990 ~tissue sarcoma, porphyria, non-Hodgkin's Iymphoma, and chloracne - EPA ~approves" other approaches to estimates of TCDD risks FIGURE 2 Timeline of significant scientific, societal, and regulatory events in- volving the dioxin TCDD. -
224 MOORE, KIMBROUGH, AND COUGH association with birth defects was discovered, since the incidence of off- spring with birth defects was similar to that seen with untreated mice (Lamb et al., 198 la,b). Another important health study was reported in 1978. In a long-term study in rats, an increased incidence of tumors was associated with TCDD exposure (Kociba et al., 1978~. These results were soon confirmed by other studies in mice and rats; currently there are well over a dozen research publications associating TCDD with the development of cancer in animals; it is generally held that TCDD "promotes," as opposed to "initiates," the expression of a carcinogenic response, probably through some form of hor- monal interaction. Epidemiological Studies Epidemiological studies of the effects of TCDD in humans are almost exclusively investigations in males. For example, there are no definitive human studies of reproductive and developmental effects in women exposed to TCDD. The one study of exposures to women in and around Seveso, Italy, after TCDD contamination, did not demonstrate an increased risk of birth defects (Mastroiacovo et al., 1988~. However, the statistical power of the finding was limited by the size of the study population. Most of the studies done to date have focused on male Vietnam veter- ans, the Air Force personnel that applied Agent Orange in Vietnam (Opera- tion Ranch Hand), or workers whose occupations resulted in exposure to TCDD. In response to veterans' concerns about birth defects, the Centers for Disease Control (CDC) undertook a case control study using data from the Metropolitan Atlanta Congenital Birth Defects Program. This CDC birth defects study was designed to determine if Vietnam veterans were fathering children with a higher incidence of birth defects. The potential risk factor being investigated was service in Vietnam, not TCDD exposure, per se, since there were no records that allowed such a distinction. The study results did not establish an association between service in Vietnam and a higher incidence of birth defects in children (Erickson et al., 1984~. The Ranch Hand studies, which are continuing, are the most thorough in breadth and frequency of medical surveillance. The study focuses on the 1,200 persons that were part of the Ranch Hand operation, ranging from pilots and crew that flew the missions to the ground personnel that serviced the aircraft and handled the storage and loading of Agent Orange. These individuals periodically receive a thorough medical examination, and fertility is followed as are incidence of disease and cause of death. Thus far, these studies show no patterns that indicate a health detriment due to TCDD exposure. Significant associations between serum dioxin - -
THE DIOXIN TCDD 225 levels and variables related to lipids (fats that are not soluble in water) were found. Since TCDD is soluble in lipids, the association may reflect a difference in the pharmacokinetics of TCDD in different subsets of the Ranch Hand population. Many individuals in the Ranch Hand study were found to have elevated TCDD levels in their blood serum (Air Force Health Study, 1991), providing objective evidence of the degree of past exposure to Agent Orange or possi- bly other herbicides containing TCDD that were used in comparatively small quantities in Vietnam. Other researchers have reported increased TCDD levels in adipose tissue (fatty tissue) in veterans who were associated with herbicide application in Vietnam (Schecter et al., 1989~. In contrast, studies of veterans of ground warfare in Vietnam have not reported increased levels of dioxin (CDC, 1988~. In the absence of an identifiable group of ground warfare veterans ex- posed to TCDD, several studies have compared the health of Vietnam veter- ans with that of other veterans. One of those studies reported a service- related increase in non-Hodgkin's lymphoma, which is unrelated to exposure to TCDD (Selected Cancers Cooperative Study Group, 1990~. This increase in incidence was higher among men in the sea-based Navy, a group that is not believed to have had exposure to either Agent Orange or TCDD. No significant adverse effects have been reported in other studies (CDC, 1988; Kahn et al., 1988; Kang et al., 1991; Cough, 1991~. Hardell and coworkers (Ericksson et al., 1981; Ericksson et al., 1990; Hardell et al., 1979; Hardell, 1981; Hardell et al., 1981) studied agricultural and forestry workers and reported a positive association between exposure to chlorinated phenols or phenoxy acid herbicides and an increased inci- dence of soft tissue sarcoma or non-Hodgkin's lymphoma. Others have failed to confirm these findings in agricultural and forestry workers (Woods, 1987~. The reports of Hardell and coworkers are most directly contradicted by an examination of TCDD levels in New Zealand herbicide sprayers. Although the New Zealand men have elevated levels of TCDD, there is no evidence of any health effects resulting from their exposure to herbicide. It was reported that dioxin levels in patients who were part of Hardell's stud- ies were similar to those in control patients who were undergoing gall blad- der surgery (Rappe et al., 1984) Another publication from the same group of Swedish investigators reported that there was no difference in the levels of TCDD or the pattern of other chlorinated dibenzodioxins and dibenzofurans in adipose samples collected from cancer (soft tissue sarcomas or lymphomas) patients classified as "exposed" to phenoxy herbicides and a group of con- trols that had no history of phenoxy acid herbicides (Nygren et al., 1986~. The authors of the New Zealand study conclude that the associations re- ported by Hardell et al. are almost certainly in error (Smith et al., 1992~. A combined mortality study of chemical workers from twelve different ~7
226 MOORE, KIMBROUGH, AND COUGH U.S. plants reported an increase in total tumors, and tumors of the lung and bladder in workers exposed to TCDD and many other chemicals (Fingerhut et al., 1991~. An increase in soft tissue sarcomas was also observed in a subset of these workers. In highly exposed workers who became ill long after exposure (20 years), the mortality for all cancers combined was 46 percent higher than expected. This increase was primarily due to a 42 percent increase in cancers of the respiratory tract. Although the increased death rate from total cancers is dominated by cancers of the respiratory tract, very little is known about smoking in the study population. A survey of smoking habits among "surviving workers at only two plants" contrib- utes no information about the smoking habits of workers who had died at all twelve plants studied and who are the subjects of this study. Furthermore, two workers died from mesothelioma, clearly showing that some workers were exposed to other causes of lung cancer, that is, asbestos. Some of the workers who developed bladder cancer were also exposed to chemicals that are considered to be carcinogenic for the human bladder. Five of the seven workers who developed soft tissue sarcomas had worked in only one of the 12 plants included in the study by Fingerhut and cowork- ers. In workers from that plant who had soft tissue sarcoma, the effects of exposure to TCDD were confounded by exposure to the known human blad- der carcinogen para-aminobiphenyl (International Agency for Research on Cancer, 1972~. None of the workers with soft tissue sarcoma were exposed only to TCDD (Collins et al., 1993~; all were also exposed to para-aminobiphenyl. These results need more in-depth study. In support of the contention of Collins et al., a study of 1,600 German chemical workers exposed to TCDD found no soft tissue sarcomas (Zober, 1990; Manz et al., 1991~. The Ger- man study did report an increase in deaths from all cancers, and respiratory cancers are a major contributor to the overall excess of cancer deaths. So far as soft tissue sarcomas are concerned, the two studies contradict each other. Another factor that may confound the interpretation of data on soft tissue sarcoma are errors that are known to occur in the (mis~classification of this tumor type. All in all, a critical but fair reading of the epidemiology studies leads to the conclusion that deaths from some cancers were elevated in chemical plant workers with unknown smoking habits who were exposed to a variety of chemicals. Among those chemicals was TCDD. Immunology Experimental animal studies have shown other toxic effects of TCDD, particularly suppression of immune system responses. Although attempts have been made to detect similar effects in humans, causal associations between TCDD exposure and altered immune status have not been made.
THE DIOXIN TCDD 227 No consistent effects have been observed in studies of Air Force personnel who applied Agent Orange (Lathrop et al., 1984; Lathrop et al., 1987; Wolfe et al., 1991), individuals in Seveso, Italy, who developed chloracne (Mocarelli et al., 1986; Tognoni and Bonaccorsi, 1982), or in a Missouri study of persons who had documented exposure to TCDD (Evans et al., 1988; Webb et al., 1989~. These three studies are of particular value since actual body burdens of TCDD could be quantified or else chloracne was present in the study participants. To date, immunotoxic effects have not been a factor in risk management decisions. REGULATORY LIMITS For several decades it has been traditional practice in the United States and the developed countries of the world to set quantitative limits on human exposure to a toxic agent by applying a numerical safety factor to a no- observed-effect-level (NOEL). The NOEL is the highest dose in the most sensitive test species at which no toxic effects are observed. In the NOEL- plus-safety-factor approach, exposures below this level are assumed not to pose a risk because one is below the threshold of toxicity. In 1976 the U.S. Environmental Protection Agency (EPA) changed this practice in estimating cancer risk in the United States on the assump- tion that there are no NOELs for carcinogens but that risk decreases with dose, reaching zero risk at zero dose. The choice of this specific risk estimation model called the no-threshold model, was strongly influenced by studies of cancer in experimental animals exposed to radiation, the best source of data available at that time. Adoption of this assumption pro- vides numerical estimates of the number of people who may be at risk of cancer in specific situations. These estimates presumably have greater utility than the older procedures, which could identify the number of people exposed above a "safe" level but could not provide the "estimated inci- dence" of disease. Risk Models The "interim" cancer assessment procedures the EPA adopted in 1976 specified the use of linear, no-threshold models. In 1980 the EPA devel- oped an upper estimate of the human cancer risk posed by TCDD using the rat bioassay data of Kociba et al., (1978~. This estimate was modestly revised in 1985. The 1985 estimate associates a dose of 0.006 picogram/ kilogram (pg/kg)/day with a maximum plausible increased cancer risk of one additional case in a million people. The EPA Risk Assessment Guidelines for Cancer, which were issued in 1986, retained the 1977 "interim" policy in that the guidelines affirmed the -
228 MOORE, KIMBROUGH, AND COUGH linearized multistage model as the default methodology for estimating can- cer risk. Numerical estimates of cancer potency derived from this model are used by EPA in various activities such as setting water quality standards and cleanup levels for hazardous waste sites, and in estimating cancer risk from effluent or ash produced by municipal waste incinerators. Other fed- eral agencies, such as the Food and Drug Administration and the Centers for Disease Control and Prevention, have also derived estimates of TCDD can- cer risk using the same basic methods but, owing to differences in other scientific assumptions such as how to adjust for differences in size of mice, rats, and humans, their estimates have produced values that permit exposure levels that are, respectively, nine and four times higher than the EPA's estimate. Other countries have not elected to apply a linear multistage model as the standard means for estimating cancer risk. Rather, these countries evaluate risks on a case-by-case basis. The practical result of this approach is that the linearized model is used for chemicals that cause genetic toxicity and the NOEL-safety factor approach is used for all other chemicals. Since TCDD is widely interpreted by scientists not to be genotoxic (Zeiger, 1989), the latter approach has been applied to TCDD data, resulting In acceptable exposure limits that are several orders of magnitude greater than those de- rived by the U.S. EPA. TCDD as an Exception Since the mid-1980s, researchers have produced data that indicate the cancer response (and other toxic effects) to TCDD exposure is mediated through binding to a specific molecule, the Ah (aryl hydrocarbon hydroxy- lase) receptor, in the cells of mammals. Research is currently focused on understanding the molecular events associated with binding, whether there are significant differences among animal species, and the effects that ex- posure to other chemicals (dioxins, dibenzofurans, polychlorinated biphe- nyls [PCBs]) may have on the binding of TCDD. In addition, there are differences in pharmacokinetics between humans and animals (Kimbrough, 1990~. A basic understanding of these processes is needed to gain an understanding of how receptor binding may be relevant to human risk assessment. Some scientists have postulated that since TCDD does not cause ge- netic damage, it cannot "initiate" the essential and irreversible first step in the carcinogenic process. TCDD "promotes" rather than "initiates" tumors. It is plausible to conclude that TCDD affects the carcinogenic process through mechanisms that have a threshold; this makes the standard linearized, multi- stage model inappropriate for estimating cancer potency. An ad hoc EPA committee in 1986 produced the opinion that the linear model was inappro 0-
THE DIOXIN TCDD 229 priate for TCDD. In 1988, a group of senior EPA scientists concluded that "reliance on the linearized model may be less appropriate for TCDD than for many other chemicals, . . . and the model may overestimate the upper bound on the risk by some unknown amount" (EPA, 1988~. The group was of the opinion that since TCDD represents less of a toxic risk than previ- ously thought, a higher exposure level could be tolerated without increased risk. The group recommended that the risk-specific-dose (the dose which may result in a one in a million increase in cancer) be raised from 0.006 to 0.1 pa/kg/day. EPA management wanted the opinion of independent experts before it acted on the recommendation and submitted the report to its Sci- ence Advisory Board for comment. The EPA Science Advisory Board established an ad hoc committee to peer review the EPA document, and it agreed that the current linearized multistage model was inappropriate and urged the use of other available models; however, it also stated that there were no new scientific data to support a modification of the value developed in the 1985 risk assessment. The Executive Committee of the EPA Science Advisory Board, in transmit- ting the peer review comments to the administrator, took the unusual step of expressing their opinion "that the existing LMS-based risk assessment . . . lacks a firm scientific foundation" (EPA, 1988~. In a separate activity, many scientists that were experts on the effects of dioxin expressed the opinion at a 1990 Banbury Conference that using traditional models for assessing cancer risk from dioxin is not appropriate; there was consensus that the known toxic effects of dioxin are receptor mediated. The Current EPA Reassessment On April 8, 1991, the EPA administrator announced that the agency would reassess the risks of exposure to TCDD and related compounds in light of significant advances that had occurred in the scientific understand- ing of the mechanisms by which dioxin becomes toxic, the health effects in animals and people, the pathways to human exposure, and the toxic effects of dioxin in the environment. The scientific reassessment of "dioxin" is notable because it does the following things: 1. It broadens the range of scientific issues to be considered from those dominated by cancer to include a wider consideration of such key issues as exposure, dose-response relationship, pharmacokinetics, mechanism of ac- tion, immunology, and reproductive and developmental toxicity. 2. It commits the EPA to the development of a biologically based, dose-response model to estimate human health risks. This responds to opinions expressed by scientists since 1986 that the traditional linear mul- tistage model was inappropriate for estimating human risk from TCDD in.. . ,.
230 MO ORE, KIMBR O UGH, AND G O UGH exposure and implements the recommendation of the EPA Science Advi- sory Board. 3. It defines a process that enhances the opportunity for the develop- ment of a broad scientific consensus. For example, it enlists the participa- tion of a number of the country's dioxin researchers in the review of each topic. A group of scientists outside the EPA reviewed and received com- ments on the draft documents and will direct the drafting of a risk charac- terization statement based on these reviews. The latter step is unprec- edented in the history of the EPA. The revised scientific chapters and the risk characterization draft will be submitted to the EPA Science Advisory Board for review and comment. 4. It proposes to assess the toxicity potential of not just TCDD but related compounds such as other dioxins, certain dibenzofurans, and se- lected polychlorinated biphenyls. These other chemicals are believed by some to exert their toxicity through a similar mechanism. Therefore, quan- titative estimates of total risk should reflect the sum of exposures to all of these chemicals. The use of this "toxic equivalency" for dioxins and dibenzofurans was reviewed by the EPA Science Advisory Board several years ago and found acceptable as an interim measure. Extending the toxic equivalency approach beyond dioxins and dibenzofurans to selected PCBs is more controversial, however, because of concerns that the data are inadequate. For example, in 1990 an EPA workshop on PCBs, principally attended by scientists, advised against adopting such a scheme for PCBs because of large data gaps (Barnes et al., 1991~. Research data collected since the 1990 workshop, some of which were developed by EPA, indicate that the equivalency values for some PCB congeners (i.e., other chemicals in the same group) are inaccurate by more than two orders of magnitude or incorrectly predict for certain toxic effects, such as cancer potency. The dioxin reassessment was originally scheduled to be completed in the latter part of 1992; current estimates are that the revised assessment will be submitted to the EPA Science Advisory Board in the spring of 1993 and, in turn, their review will be transmitted to the administrator in late 1993. It is reasonable to expect that completion of the reassessment, irrespective of its content, will have significant implications for science policy and risk management policy. THREE CASES OF RISK REGULATION To illuminate the interplay between new information on toxicity and public policy regarding regulation of risk, we explore 2,4,5-T, Agent Or- ange, and the Clean Water Act in greater detail.
THE DIOXIN TODD 231 The Pesticide 2,4,5-T The chemical 2,4,5-T was first registered for use as a herbicide in 1949 and quickly gained wide domestic use for controlling a variety of broad- leafed weeds and woody plants. The registered uses permitted its applica- tion in a wide variety of homeowner and agricultural circumstances that included croplands, forestry, and a diverse array of brush control practices, including roadsides, railroad and electrical power line rights of way, and rangeland. The data indicating TCDD's potential for causing birth defects and other forms of developmental abnormalities in laboratory studies led to the 1970 decision by the U.S. Department of Agriculture, the agency that then regu- lated pesticides in the United States, to eliminate many domestic uses of 2,4,5-T, including homeowner use and use on food crops. Continued uses in forestry, on rangeland, on rights of way, and in rice fields were permitted, resulting in the application of 6 to 7 million pounds of 2,4,5-T in 1974 (Gough, 1986, pp. 137-138~. In 1978 the Environmental Protection Agency, which by then had as- sumed responsibility for regulating pesticide use in the United States, an- nounced that it was considering cancelling all uses of 2,4,5-T and other herbicides made from trichlorophenol (Gough, 1986, p. 138~. This action was motivated by concern about the developmental toxicity reported in the original series of experimental animal studies and was confirmed and ex- tended in subsequent investigations; these included studies in rats that spanned several generations (Murray et al., 1979~. In the spring of 1979, the EPA suspended, on an emergency basis, essentially all remaining uses of 2,4,5-T and related herbicides. The sus- pension was based on the laboratory data as well as new epidemiology data that purported to correlate herbicide use in Oregon with a seasonal increase of human miscarriages. These studies, involving people residing in the Alsea, Oregon area, were reported widely in the press. At the time, EPA policymakers were persuaded to take the extraordinary action of invoking an emergency suspension based on two current facts: the Alsea study results and knowledge that the start of annual spraying of the forests was immi- nent. The policymakers were briefed by agency scientists on the adequacy of the epidemiology data and reportedly received similar opinions from a few academic scientists whose cursory opinions were solicited.) Detailed analysis of the Alsea results, which occurred during the cancellation pro- cess, led most scientists to conclude that there were major flaws in the study that made the authors' conclusions untenable. The toxic effect of concern, at that time, remained focused on developmental abnormalities, as reported in humans and demonstrated in experimental animals. In 1983, after exten- sive hearings on the merits of permanent cancellation of 2,4,5-T, the last -
232 MOORE, KIMBROUGH, AND COUGH company that held a pesticide registration withdraw from the cancellation proceedings being conducted by the EPA. As a result, a regulatory decision based on the administrative hearings (which would have included a detailed discussion of health risk) was never issued and the pesticide registrations were cancelled through established administrative processes. Agent Orange Agent Orange was the name given by the military to a pesticide formu- lation that was composed of equal parts of the herbicides 2,4,5-T and 2,4-D. At the time Agent Orange was formulated, both herbicides were registered for use in the United States. The military's evaluation of herbicides in Vietnam led to the conclusion that there was significant military benefit associated with their use. The military use of chemicals as defoliants was a politically sensitive issue because of fears that it would be viewed internationally as a form of chemical warfare. Many Americans were opposed to the war; some Ameri- can scientists were also vocal in their opposition to herbicide use. Several scientific organizations formally expressed their concern that broad use would cause significant environmental damage (Gough, 1986, pp. 54-55~. It has been estimated that up to 10 percent of the land mass of South Vietnam was sprayed with a herbicide by the military (Gough, 1986, p. 49~. Herbicide use in Vietnam began in 1962. Because of its effectiveness, Agent Orange became the preferred defoliant in 1965, accounting for 60 percent of the total herbicide used in Vietnam. The peak spraying years were 1967 to 1969, with an estimated total of 10 to 12 million gallons of Agent Orange applied (Gough, 1986, p. 51~. The data indicating that a constituent, 2,4,5- T, and the TCDD contaminant could cause birth defects in mice was cited as a major factor in first limiting use of the herbicide in 1969 to areas that were not heavily populated and, finally, ending its use in 1970. In the mid-1970s a number of veterans sought health care from the VA for psychological and organic illnesses they believed were a consequence of service in Vietnam. It was a nurse in a Chicago VA hospital who first gained public attention by asserting that many of these effects were due to exposure to herbicides used in Vietnam. Organizations representing veter- ans soon began aggressive lobbying for Vietnam veterans to receive disabil- ity compensation for a range of illnesses. A variety of committees were established to review data and advise on these issues. On the recommenda- tion of an advisory committee, the Department of Veterans Affairs (VA) currently compensates veterans that served in Vietnam if they develop chloracne or soft tissue sarcoma. Based on a decision of the Secretary of Veterans Affairs, compensation is also granted to veterans who develop non-Hodgkin's lymphoma. The VA took this action after being briefed by the Centers for ,
THE DIOXIN TCDD 233 Disease Control that there appeared to be an increase in this disease in servicemen who served in Vietnam; it does not correlate with increased Agent Orange exposure. Currently, there is a proposed regulation that would compensate veterans for peripheral neuropathy that developed within 10 years of the last date of service in Vietnam. As specified in the Agent Orange Act of 1991 (P.L. 102-4), the Institute of Medicine of the National Academy of Sciences has undertaken a review and evaluation of the available scientific evidence regarding associations be- tween disease and exposure to dioxin and other chemical compounds in herbi- cides used in Vietnam. The report is to be issued in the latter part of 1993 and will provide scientific and medical information for the Secretary of Veter- ans Affairs to consider when making determinations about compensation. Clean Water Act Section 303 of the Clean Water Act requires states to adopt water qual- ity standards to protect the public health or welfare and enhance the quality of water. States must adopt numeric criteria for all listed pollutants and submit new or revised standards, containing numeric values, to the EPA for review and approval. TCDD is a listed toxic pollutant and is a chemical for which EPA developed and published recommended criteria for human health and aquatic life. The recommended numerical values for water from which fish may be caught for human consumption is 0.014 parts per quadrillion. fIf humans also drink the water, the value is 0.013 parts per quadrillion (ppq).] Sixteen states have adopted the EPA recommended value, which is equated with a maximum increased cancer risk of one in one million. In 1990 the state of Maryland proposed a numerical value of 1.2 ppq, almost 100-fold higher than the recommended EPA value. As required by the Clean Water Act, the Maryland proposal was reviewed by EPA and ap- proved on September 12, 1990. In its letter of approval, EPA clearly ac- knowledged that there are varied choices that can be made in the develop- ment of a risk assessment. While some of these choices may differ from those selected by EPA, they are equally defensible from a scientific per- spective. In the Maryland proposal, many of the scientific assumptions were the same as those of EPA; however, where they differed, assumptions that had been used by the Food and Drug Administration were used. In addition, Maryland chose to employ a policy that would tolerate an in- creased cancer risk level of one in one hundred thousand, a decision that is clearly within the prerogative of a state under the Clean Water Act. This policy initiative was facilitated by language in the Clean Water Act that gives the states the right to develop different standards as long as they are scientifically defensible. Five additional states have received approval for numerical values similar to those adopted by Maryland. -
234 MOORE, KIMBROUGH, AND COUGH THE PUBLIC PERCEPTION OF CHEMICAL RISK Public policy is not developed in a vacuum nor is it the handmaiden of scientific information. It is made by people who are well aware of the currents and countercurrents of the social issues of the day. Specific deci- sions are made within a social context. Coincident with the development of experimental data on the toxic effects of TCDD, a series of events occurred that fostered the growing unease that the public was subject to significant chemical exposures over which they had no control and which could result in adverse effects. Consider the following examples: - · The late 1960s and early 1970s was the impact period of Rachel Carson's book, Silent Spring, which forcefully and graphically portrayed the consequence of what appeared (and probably was) to be an indiscrimi- nate use of many pesticides. · The Vietnam War was controversial; spraying chemical defoliants as part of the war was even more so. In 1970 the public learned of the presence of a mystery chemical, TCDD, in sprays that were used to defoli- ate large areas in Vietnam. Humans were exposed to these sprays; in some instances it may have been our own troops. Laboratory tests of its toxicity were not concluded until after an extensive period of spraying, a fact that seemed to fit the pattern of indiscriminate use described by Rachel Carson. · In 1974, studies in Missouri revealed that waste oil, applied to dirt roads and horse arenas to suppress dust, contained TCDD and was respon- sible for sickness and death in horses and possibly accounted for illness in two youngsters (Carter et al., 1975~. In 1982 an entire town, Times Beach, Missouri, was purchased by the state and federal government, who stated the action was necessary as a result of TCDD contamination that occurred through the application of waste oil to unpaved streets. · In the summer of 1976, thousands of people in the town of Seveso, Italy, were contaminated with TCDD when an overpressured reaction vessel in a chemical plant vented its contents to the outside air. Many hundreds of animals became sick, some died (the effects were only partially due to TCDD; many were caused by other chemicals, a fact that was not grasped, and therefore poorly reported by the media). Portions of the town adjacent to the plant were evacuated. . The 1979 Alsea, Oregon, episode, widely reported on television, brought the issue of aerial spraying "home" in that it was U.S. forests that were sprayed, domestic watersheds that were possibly contaminated with TCDD, and American women who feared that their miscarriages resulted from the spraying. Other uses of 2,4,5-T caused some to wonder if contamination of meat and milk could occur. · In the early 1980s, companies that had manufactured Agent Orange
THE DIOXIN TCDD 235 agreed to a settlement of a class action suit filed in the federal court. The settlement created a $180-million fund to be shared by Vietnam veterans. From the perspective of the general public, it appeared that chemical com- panies paid a large sum to settle an issue (Agent Orange) while staunchly maintaining that the chemical did not cause health effects (i.e., the axiom that actions speak louder than words). DISCUSSION The saga of TCDD portrays the policy process accommodating scien- tific and medical information in the context of broader social issues. In some instances, scientific data on adverse effects were persuasive for policy- makers and in other instances scientific doubt or uncertainty as to adverse effects resulted in hesitancy to establish or change policy. Courtney and colleagues reported in 1969 and 1970 that TCDD could cause birth defects in animals. These results had a profound impact on management policy and regulatory decisions. The noteworthy result, from a scientific perspective, was not the nature of the developmental abnormality but the exceedingly low doses of TCDD that could lead to this developmen- tal effect. Thus began the fascination of scientists for studying how TCDD could elicit biological effects that range from birth defects to cancer, immu- nosuppression, and death. TCDD also causes a potent and persistent stimu- lation of important enzymes. The unpublished results of the laboratory tests, showing that the dioxin in 2,4,5-T caused teratogenic effects, were leaked to the press, and soon after a congressional hearing was held. The focus of the hearing was con- cern that the dioxin in Agent Orange was causing birth defects in Vietnam- ese civilians. The hearing was well covered by the press, and many mem- bers of the public were introduced to information on dioxin and its toxic effects at that time. The toxicity data indicating potential for developmental abnormalities led to decisions to restrict and subsequently suspend the use of Agent Or- ange in 1970. The rapid pace of the decision-making process was undoubt- edly influenced by the social turmoil associated with the war itself and the belief of some people that herbicides were an insidious form of chemical warfare. The Agent Orange issue resulted in social pressure to review the broad range of domestic uses of 2,4,5-T in the United States. Again, the scientific data drove the decision to restrict its use but the rapid pace at which the pesticide registrations on many products were revised or cancelled by the Department of Agriculture in 1970 may have been influenced by agricul- tural interests and pesticide manufacturers who preferred that the issue be resolved before pesticide regulation was transferred to the newly created -
236 MOORE, KIMBROUGH, AND COUGH EPA. While use of these herbicides by the general population was essen- tially eliminated, as were many agricultural uses, other agricultural uses that were believed not to involve significant human exposures were re- tained.2 Several years later, scientific data on birth defects again influenced congressional and legal decisions. Specifically, the negative results from the CDC birth defects study persuaded Congress not to compensate Vietnam veterans for children born with birth defects. In the Agent Orange class action suit, brought on behalf of Vietnam veterans against manufacturers, the presiding judge, Judge Jack Weinstein, in summing up the evidence stated, "no laboratory nor epidemiologic evidence exists at this time that is sufficient to link deaths or birth defects to parental exposures to herbicides while serving in Vietnam" (Gough, 1986, pp. 115, 117~. The degree to which scientific information influenced the Veterans Ad- ministration and the Congress on the issue of carcinogenicity is not clear. The mandate of the Department of Veterans Affairs is to provide care and compensate for illnesses that are plausibly service connected, a standard that is less rigorous than establishing causality as a result of exposure to a herbicide containing TCDD. Some argued that a causal link needed to be established between the illness and TCDD (Agent Orange). Others postu- lated that providing medical service, implying that an illness may be related to Agent Orange, was a policy without a scientific base. While scientific certainty is not a prerequisite to the VA policy decision-making process, "granting of medical service" fosters the perception that links have been clearly established.3 ~1 To the public, the issue of estimating the cancer risks associated with exposure to TCDD remains unsettled. Cancer risk was the dominant health effect cited in the decision to purchase Times Beach, Missouri, an action that is regarded by many as a politically expedient decision of the moment in which TCDD health concerns were subsequently invoked as a rationale. At any rate, this action provides a dramatic contrast to the decision to reoccupy Seveso, Italy, when reduced TCDD levels in much of the city were attained. The levels of TCDD that posed a health risk for Times Beach were estimated using linear multistage models. Thus, these levels were several orders of magnitude lower than if the alternative model, using the NOEL with a safety factor, had been used. favored by the Europeans. In some respects the central issue regarding the estimation of TCDD cancer risk may not be TCDD per se, but EPA's adherence to a method of estimating cancer risks that provides results far more restrictive than re- quired by its current cancer guidelines. The EPA still uses TCDD risk estimates derived from the linearized model. While it did propose a modest revision of the estimate in 1988, it seemingly was thwarted by a scientific Use of the latter model is _ .
THE DIOXIN TCDD 237 peer review that stated there was no new scientific evidence to support a change from the current estimates. In retrospect, the peer review group should have been assigned the additional task of determining if the scien- tific basis for the existing EPA position was any stronger or more compel- ling than the position that EPA had proposed to adopt. At the same time the reviewers stated that there were no new data to support change, they were critical of the methodology used to develop the existing estimates of risk! There was ample language in the peer review report to permit a policy determination like that which was proposed, that is, increase the exposure level from 0.006 pa/kg/day to 0.1 pa/kg/day. In response to the peer review, senior policy officials maintained the current standard. A number of months later, the same officials directed the start of a broad reassessment of health and environmental risks associated with exposure to dioxins. That process, which is still going on, gives a significant degree of flexibility in interpretation and judgment to expert scientists drawn from a broad array of interests across the country. This approach contrasts with the traditional practice in which the initial assess- ment is developed by agency risk assessors, with revisions based on public comment. Observing that a consensus had emerged among expert scientists at a recent conference regarding initial biological events following TCDD exposure, these EPA officials believed that it was an opportune time to revisit current risk assessment practices. The continuing collegial approach encourages a wider dialogue and has the potential to better reflect a broad- based scientific consensus. During this same period, regulatory action under the Clean Water Act reflected an explicit acknowledgment that alternative assumptions and prac- tices can yield risk estimates different from those espoused by EPA that are of equal scientific plausibility. Constraints in the law may have led the EPA to approve the more permissive state TCDD water quality standards. A broader sentiment supports the view that EPA policymakers seized an op- portunity to endorse a mechanism for setting TCDD exposure levels that is less constraining than the existing agency procedure. By doing so, they approved values that are closer to those derived by other federal agencies. This action under the Clean Water Act, as well as prompting the dioxin reassessment process described above, suggests that the EPA's rigid institu- tional commitment to a narrow modeling approach for assessing cancer risks may have been weakened. If so, EPA may become more receptive to review and use of newer knowledge. TCDD also appears to be the catalyst for fundamental changes in a number of procedures used to examine scientific data in order to assess health risk and express it quantitatively. The EPA reassessment process, as well as the Institute of Medicine report, which is to be submitted to the VA, is breaking new ground in addressing a long-standing need: a process in , ..
238 MOORE, KIMBROUGH, AND COUGH which scientific judgments are periodically reviewed and either reaffirmed or revised as necessary. This is not a revolutionary concept; periodic re- view of national ambient air quality standards was an implicit requirement in the early Clean Air Acts. The fact that this is now occurring may repre- sent a belated evolution in actual practices. A policy judgment about a chemical risk almost always represents deci- sion making in the face of uncertainty. The principal elements leading to uncertainty are a paucity of data that describe the effect; lack of data that characterize the nature and degree of exposure; and little understanding of the physiological processes involved in the observed effect. In such circum- stances, assumptions are made and uncertainty factors incorporated which, in principle, should be replaced with actual data when they become avail- able. In practice, however, the system does this only grudgingly, after concerted effort. It is to be hoped that a process that rewards greater "cer- tainty" of scientific judgment would serve as an incentive for the develop- ment of better information. The dioxin reassessments also venture into issues of chemical additivity and antagonism, and force risk assessments to be broadened to health ef- fects that transcend cancer. They also present an opportunity (some would term it an obligation) to define a more significant role and utility for data on humans. In some respects an attitude seems to have evolved in risk assess- ment in which laboratory data and model estimates reign supreme. While animal data are the legitimate backbone of preventive toxicology, human data and medical judgment merit a more prominent role. For example, sound human data exist on immune response as a function of body burden of TCDD. There are also experimental data on animals. In instances where both animal and human data yield different estimates of risk, should accept- able levels of exposure be derived from the human data or the animal data? For a number of dioxin risk considerations, this choice will be the crucial . . pO lCy Issue. From a management and policy perspective, the issues that need to be addressed sometimes are clearer than either science or risk assessment per- ceives them to be. For example, the animal data indicate that TCDD is a carcinogen. The long-established default assumption has been to treat TCDD as if it were a human carcinogen based on these data. With our existing level of knowledge, this assumption is unlikely to change. A debate contin- ues on the "proper" interpretation of human epidemiology data regarding cancer. From a policy perspective, this is not a major issue; the debate is not going to yield a definitive answer; thus, based on animal data, TCDD will continue to be viewed as a human carcinogen. However, of central importance for both regulatory and VA policy is scientific judgment regard- ing the dose that may increase the risk of cancer. Thus the extensive epide- miology data on chemical workers assembled by Fingerhut et al. (1991) I
THE DIOXIN TCDD 239 may be useful for quantitative estimates. Even if one assumes (and many vehemently oppose this as an assumption) that the relative rates of cancer are related to TCDD exposure, these rates can only be viewed as a relatively weak (low) response in a population where exposure was probably a thou- sandfold higher than the highest levels occurring today. That is, it is a weak response relative to the high estimate of potency derived from models of animal data. In a more pedestrian way, it certainly could affect the state- ment commonly uttered that "EPA considers dioxin to be one of the most potent carcinogens studied." A final area in which risk assessment is changing as a consequence of TCDD is the tacit acknowledgment that data that identify mechanisms by which adverse effects develop can be very helpful in developing more meaning- ful scientific judgments of risk. Exploring the development and use of biologically based risk assessment models, which is a priority element of the EPA reassessment, is an important change in the philosophic approach to evaluating the risk of cancer. Such changes in the types of scientific data used to assess chemical risk are vital if scientific knowledge is to more accurately inform and enlighten those responsible for social policy. NOTES 1. The emergency suspension withstood legal challenge in Federal District Court. Under emergency suspension, sale and use ceases immediately; in essence, a product is off the market while the process that leads to a decision to permanently cancel, or otherwise modify condi- tions of use, is under way. The cancellation process entails the preparation and presentation of documents that contain the data and other information that form the bases for the cancellation. There is an opportunity for review and comment on this document, including administative hearings, before a final decision is issued by the EPA. 2. The statute under which pesticides are registered mandates that regulatory decisions be based on a risk-benefit analysis on the grounds that the mere presence of risk alone is insuffi- cient to prohibit use. The risks of use must be weighed against the benefits to be realized from use. 3. The process of reconciling policies with current scientific knowledge about TCDD con- tinues, as exemplified by the requirements of the Agent Orange Act of 1991, described earlier. Legislation calling for a National Academy of Sciences report on current scientific understand- ing underscores the continuing unease and uncertainty regarding the scientific facts as well as the desire that the data and scientific judgments may guide a VA policy that fairly responds to the health concerns and afflictions of Vietnam veterans. REFERENCES Air Force Health Study. 1991. Serum dioxin analysis of 1987 examination results. Epidemi- ology Research Division, Brooks Air Force Base, Texas. Barnes, D., A. Alford-Stevens, L. Birnbaum, F. W. Kutz, W. Wood, and D. Patton. 1991. Toxicity equivalency factors for PCBs? Quality Assurance: Good Practice Regulation and Law 1(1):70-81. Carter, C. D., R. D. Kimbrough, J. A. Liddle, R. E. Cline, M. M. Zack, and W. F. Barthel.
240 MOORE, KIMBROUGH, AND COUGH 1975. Tetrachlorodibenzodioxin: An accidental poisoning episode in horse arenas. Sci- ence 188:738-740. .. Centers for Disease Control Veterans Health Study. 1988. Serum 2,3,7,8-tetrachlorodizenzo- p-dioxin levels in U.S. Army Vietnam-era veterans. JAMA 260:1249-1254. Collins, J. J., M. E. Strauss, G. J. Levinskas, and P. R. Conner. 1993. The mortality experi- ence of workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a trichlorophenol pro- cess accident. Epidemiology 4:7-13. Courtney, K. D., D. W. Gaylor, M. D. Hogan, H. L. Falk, R. R. Bates, and I. Mitchell 1970. Teratogenic evaluation of 2,4,5-T. Science 168:864-866. Courtney, K. D., and J. A. Moore. 1971. Teratology studies with 2,4,5-trichlorophenoxyacetic acid and 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicology and Applied Pharmacology 20:396- 403. Ericlcson, J. D., J. Mulinare, P. W. McClain, T. G. Fitch, L. M. James, A. B. McClearn, and M. J. Adams, Jr. 1984. Vietnam veterans' risks for fathering babies with birth defects. JAMA 252:903-937. Ericksson, M., L. Hardell, and H. O. Adami. 1990. Exposure to dioxins as a risk factor for soft tissue sarcoma: A population based case control study. Journal of the National Cancer Institute 82:486-490. Ericksson, M., L. Hardell, N. O. Berg, T. Molter, and O. Axelson. 1981. Soft-tissue sarcomas and exposure to chemical substances. A case-referent study. British Journal of Industrial Medicine 38:27-33. Evans, R. G., K. B. Webb, A. P. Knutsen, S. T. Roodman, D. W. Roberts, J. R. Bagby, W. A. Garrett Jr., and J. S. Andrews, Jr. 1988. A medical follow-up of the health effects of long term exposure to 2,3,7,8-tetrachlorodibenzo p-dioxin. Archives of Environmental Health 43:273-278. Fingerhut, M. A., W. E. Halperin, D. A. Marlow, L. A. Piacitelli, P. A. Honchar, M. H. Sweeney, A. L. Greife, P. A. Dill, K. Steenland, and A. J. Suruda. 1991. Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. New England Jour- nal of Medicine 324:212-218. Gough, M. 1986. Dioxin, Agent Orange: The Facts. New York: Plenum Press. Gough, M. 1988. Science policy choices and the estimation of cancer risk associated with exposure to TCDD. Risk Analysis 8:337. Gough, M. 1991. Editorial: Agent Orange: Exposure and Policy. American Journal of Public Health 81 :289-290. Hardell, J. L. 1981. On the relation of soft tissue sarcoma, malignant lymphoma, and colon cancer in phenoxy acids, chlorophenols and other agents. Scandinavian Journal of Work Environment and Health 7:119-130. Hardell, L., M. Ericksson, P. Lenner, and E. Lundgren. 1981. Malignant lymphoma and exposure to chemicals, especially organic solvents, chlorophenols and phenoxy acids: A case control study. Scandinavian Journal of Work Environment and Health 43:169-176. Hardell, L. and A. Sandstrom. 1979. Case-control study: Soft-tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols. British Journal of Cancer 39:711-717. International Agency for Research on Cancer. 1972. IARC Monographs on the Evaluation off Carcinogenic Risk of Chemicals to Man, Volume 1. International Agency for Research on Cancer, World Health Organization, Lyon. Kahn, P. C., M. Gochfeld, M. Nygren, M. Hansson, C. Rappe, H. Velez, T. Ghent-Guenther, and W. P. Wilson. 1988. Dioxins and dibenzofurans in blood and adipose tissue of Agent Orange-exposed veterans and matched controls. JAMA 259:1661-1667. Kang, H. K., K. K. Watanabe, J. Breen, J. Remmers, M. G. Conomos, J. Stanley, and M. Flicker. 1991. Dioxins and dibenzofurans in adipose tissue of U.S. Vietnam veterans and controls. American Journal of Public Health 81:344-349. - -
THE DIOXIN TODD 241 Kimbrough, R. D. 1990. How toxic is 2,3,7,8-tetrachlorodibenzodioxin to humans? Journal of Toxicology and Environmental Health 30:261-271. Kociba, R. J., D. G. Keyes, J. E. Beyer, R. M. Carreon, C. E. Wade, D. A. Dittenber, R. P. Kalnins, L. E. Frauson, C. N. Park, S. D. Barnard, R. A. Hummel, and C. G. Humiston. 1978. Results of a two-year chronic toxicity and oncogenicity study of 2,3,7,8- tetrachlorodibenzo-p-diox~n in rats. Toxicology and Applied Pharmacology 46:279-303. Lamb, J. C. 4th, T. A. Marks, B. C. Gladen, J. W. Allen, and J. A. Moore. 1981a. Male fertility, sister chromatic exchange, and germ cell toxicity following exposure to mixtures of chlorinated phenoxy acids containing 2,3,7,8-tetrachlorodibenzo-p-dioxin. Journal of Toxicology and Environmental Health 8:825-834. Lamb, J. C. 4th, J. A. Moore, T. A. Marks, and J. K. Haseman. 1981b. Developmental and viability of offspring of male mice treated with chlorinated phenoxy acids and 2,3,7,8- tetrachlorodibenzo-p-dioxin. Journal of Toxicology and Environmental Health 8:835- 844. Lathrop, G. D., W. H. Wolfe, R. A. Albanese, and P. M. Moynahan. 1984 (unpublished). An epidemiologic investigation of health effects in Air Force personnel following exposure to herbicides. Baseline morbidity study results. U.S. Air Force School of Aerospace Medicine, Aerospace Medical Division, Brooks Air Force Base, Texas. Lathrop, G. D., W. H. Wolfe, S. G. Machado, J. E. Michalik, and T. G. Karrison. 1987 (unpublished). An epidemiologic investigation of health effects in Air Force personnel following exposure to herbicides. First follow-up examination results. U.S. Air Force School of Aerospace Medicine, Aerospace Medical Division, Brooks Air Force Base, Texas. Manz, A., J. Berger, J. W. Dwyer, D. Flesch-Janys, S. Nagel, and H. Waltsgott. 1991. Cancer mortality among workers in chemical plant contaminated with dioxin. The Lancet 338:959- 964. Mastroiacovo, P., A. Spagnolo. E. Marni, L. Meazza, R. Bertollini, G. Segni, and C. Borgna- Pignatti. 1988. Birth defects in the Seveso area after TCDD contamination. JAMA 259:1668-1672. Mocarelli, P., A. Marocchi, P. Brambilla, P. Gerthous, D. S. Young, and N. Mantel. 1986. Clinical laboratory manifestations of exposure to dioxin in children: A six year study of the effects of an enviromental disaster near Seveso, Italy. JAMA 256:2687-2695. Murray, F. J., F. A. Smith, K. D. Nitschke, C. G. Humiston' R. J. Kociba, and B. A. Schwetz. 1979. Three-generation reproduction study of rats given 2,3,7,8-tetrachlorodibenzo-p- dioxin in the diet. Toxicology and Applied Pharmacology 50:241-252. Nygren, M., C. Rappe, G. Lindstrom, M. Hansson, P. A. Bergqvist, S. Marklund, L. Domellof, L. Hardell, and M. Olsson. 1986. Identification of 2,3,7,8-substituted polychlorinated dioxins and dibenz~furans in environmental and human samples. Chlorinated Dioxins and Dibenzofurans in Perspective. Chelsea, Mich.: Lewis Publishers. Rappe, C., P. Bergqvist, M. Hansson, L. Kjeller, G. Lindstrom, S. Marklund, and M. Nygren. 1984. Chemistry and analysis of polychlorinated dioxins and dibenzofurans in biological samples. Banbury Report. Schecter, A., J. D. Constable, J. V. Bangert, H. Tong, S. Arghestani, S. Monson, and M. Gross. 1989. Elevated body burdens of 2,3,7,8-tetrachlorodibenzodioxin in adipose tissue of United States Vietnam Veterans. Chemosphere 18:431-433. Schultz, K. H. 1957. Klinische und experimentelle Untersuchungen zur Atiologie der Chlorakne. Archiv fuer Klinische und Experimentelle Dermatologie 206:589-596. Selected Cancers Cooperative Study Group. 1990. The association of selected cancers with service in the U.S. Military in Vietnam. I. Non-Hodgkin's Lymphoma. Archives of Internal Medicine 150:2473-2483. Smith, A. H., D. G. J. Patterson Jr., M. L. Warner, R. MacKenzie, and L. L. Needham. 1992. - _ r~
242 MOORE, KIMBROUGH, AND COUGH Serum 2,3,7,8-tetrachlorodibenzo-p-dioxin levels of New Zealand pesticide applicators and their implication for cancer hypotheses. Journal of the National Cancer Institute 84:104-108. Sparschu, G. L., F. L. Dunn, R. W. Lisowe, and V. K. Rowe. 1971. Study of the effects of high levels of 2,4,5-trichlorophenoxyacetic acid on foetal development in the rat. Food and Cosmetics Toxicology 9:527-530. Tognoni, G., and A. Bonaccorsi. 1982. Epidemiological problems with TCDD (a critical view). Drug Metabolism Reviews 13 :447~69. U.S. Environmental Protection Agency. 1988. A cancer risk-specific dose estimate for 2,3,7,8- TCDD. External review draft. EPA/600/6-88/007. Webb, K. B., R. G. Evans, and A. P. Knutsen. 1989. Medical evaluation of subjects with known body levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Journal of Toxicology and Environmental Health 28:183-193. Wolfe, W. H., J. E. Michalek, and J. C. Miner. 1991. Air Force Health Study. An epidemiologic investigation of health effects in Air Force personnel following exposure to herbicides. Serum Dioxin Analysis of 1987 Examination Results. Epidemiology Research Division, Armstrong Laboratory, Human Systems Division (AFSC), Brooks Air Force Base, Texas. Woods, J. S., S. L. Polissar, R. K. Severson, L. S. Heuser, and B. G. Kulander. 1987. Soft tissue sarcoma and non-Hodgkin's lymphoma in relation to phenoxy herbicide and chlori- nated phenol exposure in Western Washington. Journal of the National Cancer Institute 78:899-910. Zeiger, E. 1989. Genetic toxicity. Pp. 227-238 in Halogenated Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins and Related Products, 2nd ea., R. D. Kimbrough and A. A. Jensen, eds. New York: Elsevier. Zober, A., P. Messerer, and P. Huber. 1990. Thirty-four-year mortality follow-up of BASE employees exposed to 2,3,7,8-TCDD after the 1953 accident. International Archives of Occupational and Environmental Health 62:139-157.