PETTITIONED PUBLIC HEALTH ASSESSMENT
KOPPERS COMPANY, INCORPORATED (OROVILLE PLANT)
[a/k/a KOPPERS (OROVILLE)]
OROVILLE, BUTTE COUNTY, CALIFORNIA
Dioxin-like Compounds: Definitions, Sources, Exposures, Effects, and TEFs
1 Dioxins/Furans-An Overview
1.1 What are Dioxins/Furans?
Polychlorinated dibenzo-p-dioxins (PCDDs) and -furans (PCDFs) are two related classes of chlorinated organic compounds. They have similar core structures that can be visualized as two 6-sided benzene rings connected by two oxygen bridges in PCDDs and one in PCDFs; the second bridge in PCDFs is a carbon-carbon bond (Figure 1). There are 8 different positions on a PCDD molecule and 10 on a PCDF molecule, that can be occupied by a chlorine atom or other substituent. This makes possible the existence of 75 individual variations or "congeners" of PCDDs and 135 of PCDFs. The only difference between these various congeners of PCDDs and PCDFs is the specific number and location of the chlorine atoms in each. Different congeners that share the same number of chlorine atoms, but at different locations, are referred to as isomers. Groups of isomers that contain 1, 2, 3, 4, 5, 6, 7, or 8 chlorine atoms are called mono-, di-, tri-, tetra-, penta-, hexa-, hepta-and octa-chlorinated dioxins/furans, respectively (ATSDR, 1998).
The relative toxicity or potency of various PCDDs and PCDFs is strongly influenced by the number and position of the chlorine atoms in the molecule. The most toxic dioxin, and the most extensively studied, is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). More highly chlorinated (i.e., penta- through octa-) PCDDs/PCDFs that also have chlorine atoms at the (lateral) 2, 3, 7 and 8 positions (among others) are often described as "dioxin-like compounds" in recognition of the possibility that they may share, to some extent, the established toxicities of 2,3,7,8-TCDD (ATSDR, 1998).
1.2 What are the Sources of Dioxin Contamination and Exposure?
PCDDs and PCDFs are not produced deliberately; they are unwanted byproducts that can, under special conditions, be formed during combustion and certain industrial processes. Dioxins are formed as a contaminant during the manufacture of certain chlorinated organic chemicals (e.g., pentachlorophenol [PCP] and certain herbicides). However, due to significant refinements in the manufacturing process, emissions from the chemical industry no longer represent a major source of dioxins nationwide. Today, in the U.S., older medical and municipal incinerators constitute one of the major remaining sources of dioxins still being released into the environment. Very small amounts of dioxins are also formed during the chlorine bleaching process and released into pulp and paper mill wastewater. Today, however, these emissions constitute a minor source of human exposure to dioxin, accounting on average for about 2% of daily intake (ATSDR 1998).
Estimates of average daily background exposures in the U.S. general population are 0.3 - 0.6 pg TCDD/kg/day and about 1 pg /kg/day of dioxin-like PCDDs and PCDFs (USEPA, 1994a; ATSDR 1998). (A picogram or pg is 0.000000001 or one trillionth of a gram.) Over 90% of that exposure comes from eating meat, fish, and dairy products contaminated with residues that first entered the food chain many years ago (ATSDR, 1998). Dioxins are very stable, highly lipophillic ("fat loving") compounds that, depending on the congener and the species, may also be relatively resistant to metabolism. As a result, dioxins have a strong tendency to bioaccumulate in fat, bind strongly to soils and sediments high in organic content, and persist in the environment for many years. Nevertheless, average concentrations of dioxins in biological and environmental samples have been declining since the 1970s and continue to do so (USEPA 2000a,c). Due to the existence of natural sources, however, dioxins will never disappear completely from the environment.
Before analytical techniques were sensitive enough to demonstrate otherwise, it was commonly thought that dioxins were produced exclusively as a man-made byproduct of industrial activities. It is now known, however, that dioxins pre-date not only the industrial revolution but the human race itself; they have recently been detected in 30 million-year-old clay deposits (Hayward et al. 1999; Winters, 2000). Dioxins are produced by natural, as well as anthropogenic, combustion processes, including forest fires. They are generated in very small amounts during the combustion of almost any organic material. The concentration of chlorine (which is always present in excess) is not a limiting factor (Rigo et al., 1996; Lemieux, 2000). Thus, contrary to popular belief, the burning of materials containing polyvinyl chloride (PVC) will not necessarily produce any more dioxin than does the burning of other organic materials. The rural burning of backyard trash, for example, actually produces more dioxin nationwide than does PVC manufacture (USEPA 1998;Winters, 2000).
2 Potential Adverse Health Effects of Dioxins
Some natural substances like botulinum toxin are more toxic than dioxin, but TCDD produces adverse health effects in laboratory animals at lower concentrations than any other man-made chemical. Less than one millionth of a gram per kilogram body weight (1 ug/kg) can slowly kill a guinea pig or give cancer to a rat. However, even these tiny doses are up to a million times higher than those that typically occur outside the laboratory, today. Average background doses in the U.S. and Europe are in the range of 1 trillionth of a gram per kilogram of body weight per day (0.000000000001 grams/kg/day or 1 pg/kg/day). During the Vietnam War, the herbicide formulation Agent Orange was contaminated with parts per million of dioxin (TCDD). Today, concentrations of dioxin in highly contaminated soil are measured in the low parts per billion (ppb). In food, dioxins occur in parts per trillion (ppt), in water, in parts per quadrillion (ppq), and in air, in parts per quintillion (ppqt). Each of these units of measure is 1,000 times smaller than the previous one. (See Appendix G for the definitions of standard units of concentration.) A millionth (10-6) of a gram is called a microgram (µg), a billionth (10-9) of a gram is a nanogram (ng), a trillionth (10-12) is a picogram (pg), a quadrillionth (10-15) is a femtogram , and a quintillionth (10-18) of a gram is called an attogram.. There are 28.35 grams in one ounce. (In subsequent sections, doses will generally be converted to pg/kg/day to facilitate comparison to an average human background exposure of 1 pg/kg/day.)
2.1 Animal Effects
Relatively little is known about the adverse health effects of non-TCDD dioxins, but the most toxic congener, 2,3,7,8-TCDD, is one of the most extensively studied of all known environmental toxins. Wherever sufficiently high doses of TCDD have been administered, a variety of effects have been observed in almost every animal species tested. Observed effects in animals include death, weight loss, liver toxicity, immune suppression, reproductive impairments, birth defects, and cancer (ATSDR 1998). The doses of dioxin required to produce these adverse health effects in animals vary enormously with species, as well as with strain, sex, tissue, and duration of exposure. For example, reported LD50 values for TCDD - an LD50 is the dose of a substance required to kill 50% of the exposed test animals - vary from 0.6 ug/kg (600,000 pg/kg) in male Hartley guinea pigs to 5,051 ug/kg (5,000,000,000 pg/kg) in Syrian hamsters. This represents more than an 8,000-fold difference between two species of rodent that are much more closely related to one another than either is to humans (ATSDR 1998).
Virtually all known chronic, intermediate, and acute effects levels for TCDD range upward from 100, 1,000 and 100,000 picograms per kilogram body weight per day (pg/kg/day), respectively. For non-TCDD dioxins, known effect levels in animals exceed a million pg/kg/day or 1 ug/kg/day (ATSDR 1998) In absolute terms, these effect levels are among the lowest recorded for any man-made toxic substance, which accounts for dioxin’‘s reputation as "the most toxic man-made substance known." However, relative to the current potential for environmental exposure in humans, these levels are actually quite high.
ATSDR’‘s chronic Minimum Risk Levels (MRL’‘s) are estimates of daily doses that would not be associated with any detrimental non-cancer effects over a lifetime of exposure. Most are based on animal effects and the application of conservative safety factors. For example, ATSDR’‘s chronic MRL for TCDD of 1 pg/kg/day, which approximates average background exposures in the U.S. several years ago, is based on less serious effects on social behavior in monkey offspring and a safety factor of 90 (ATSDR, 1998).
2.2 Human Effects
Humans appear to be 10 to 100 times less sensitive to the effects of dioxin than are laboratory rats and mice (Kimbrough, 1992; Aylward et al., 1998). It is generally assumed by regulatory agencies that sufficiently high dioxin exposures would produce in humans most, if not all, of the adverse health effects seen in laboratory animals. The validity of this assumption, however, has not been confirmed by the highest human exposures recorded to date, and, due to the elimination of most major sources of dioxin emissions, it is extremely unlikely that even higher exposures will ever occur in the future. No human fatality directly attributable to dioxin exposure has ever been recorded. More than 100 epidemiological studies have looked at the potential effects of dioxin and dioxin-contaminated herbicides in humans, including highly exposed workers in the chemical industry, soldiers exposed to Agent Orange in Vietnam, and persons accidentally exposed to TCDD at Seveso, Italy and at Times Beach, Missouri (Institute of Medicine 1994 ). A weight of evidence review of all of these studies indicates that chloracne is the only adverse health effect for which there is unequivocal evidence of a causal link with dioxin exposure in humans (Gotts, 1993, pg. 176; DeVito et al., 1995). Although not life-threatening, chloracne is a serious, potentially disfiguring skin eruption associated with unusually high exposures to dioxin, especially those that occurred prior to the 1980s as a result of occupational or environmental accidents.
The International Agency for Research on Cancer (IARC) recently reclassified TCDD from a "possible" to a "known" human carcinogen (IARC 1997). That classification, which ordinarily requires "sufficient" evidence in humans, was based in this case on "limited" evidence in humans, "sufficient" evidence in animals, and "supporting information" interpreted as suggesting a common mechanism of action for TCDD in various species, including humans. However, the actual mechanism by which TCDD induces adverse effects is still essentially unknown (USEPA, 1989; USEPA 2000a,b). The assumption of a common mechanism based on binding to a common cellular macromolecule (the Ah-receptor) is the first among several assumptions and inferences of uncertain scientific validity that serve as the logical foundation of the interim toxicity equivalency factor (TEF) approach (USEPA, 1989; USEPA 1994a ; USEPA 2000a,b). (See Section 3 of this appendix.)
EPA has recently announced its own intention to re-classify TCDD as "carcinogenic to humans" and other dioxin-like compounds as "likely" human carcinogens (USEPA 2000a). Like IARC, EPA bases its re-characterization of TCDD as "carcinogenic to humans" on extrapolations from animal data, hypotheses concerning the role of the Ah receptor in dioxin’‘s mode of action, and purportedly equivalent "body burdens" (on a TEQ basis) in animal and human populations with cancer (USEPA 2000b, pg 87). At the same time, however, EPA acknowledges that the data from epidemiological studies of cancer in exposed humans "do not confirm a causal relationship between exposure to dioxin and increased cancer incidence" (EPA 2000b, pg 86).
Although some data are suggestive of an association between dioxins and cancer in humans, the studies do not support any firm conclusions. Even the best studies of the most heavily exposed occupational cohorts of the past have provided only inconsistent and inconclusive evidence that TCDD might cause cancer in humans (Zober et al., 1990; Manz et al., 1991; Fingerhut et al., 1991; Steenland et al., 1999). Generally, the observed excess risks for all cancers combined and specific types of cancer were relatively small - standard mortality ratios were generally less than 2.0 - and were seen only in the most highly exposed group with estimated 2,3,7,8-TCDD exposures 100-1000 times those seen in the general population. In addition, affected individuals tended to have exposure to other carcinogenic substances besides TCDD. In the absence of controls for such confounding exposures, the observed cancer excesses could not with any confidence be attributed to TCDD alone. In one of the largest and best-conducted of these occupational studies, observed increases in lung cancer and all cancers combined (42% and 15%, respectively) became statistically insignificant when attempts were made to control for smoking (Fingerhut et al., 1991).
The highest short-term TCDD exposures ever recorded occurred in 1976 at Seveso, Italy (Zone A), where mean exposures in children were estimated to be over 3 million pg/kg (Gough, 1994, pg 251). These children were, therefore, exposed in a single day to much more TCDD than an average U.S. citizen would be exposed to in a lifetime. The highest blood level of TCDD ever recorded (56,000 ppt) was measured in a female child at Seveso just days after the accident. Yet, with the sole exception of chloracne, follow-up studies of the Seveso cohort have firmly established no consistent or unusual pattern, either for the frequency or type of outcome, attributable to TCDD exposure (Bertazzi and Domenico 1994). Ten to fifteen years after the 1976 accident, there was an increase in some types of cancer, a deficit in others, and an overall decrease in expected cancer rates, especially among those with the highest exposure (Bertazzi et al., 1989, 1997). Thus, if dioxin is a human carcinogen, it must be a very weak one at doses that are realistically achievable outside the laboratory. In any case, the cancer risks that might be associated with dioxin exposures today would be undetectable and, hence, unverifiable.
A number of studies have reported measured differences in various developmental outcomes (e.g., levels of thyroxin, certain hormones, liver enzymes, and vitamin K; neurological endpoints; white blood cell counts and other immunological markers) in the breast-fed infants of mothers whose milk contained elevated levels of dioxin and dioxin-like compounds (Koppe, et al., 1991;Pluim, et al., 1993, 94a, 94b; Koopman-Esseboom et al., 1994; Huisman et al., 1995; Weisglas-Kuperus et al., 1995). However, the detected differences generally involve subtle and inconsistent biochemical changes that were well within the range of normal variation and are of no known clinical significance. A caveat common to virtually all such studies is that any hypothetical risks that may be associated with the consumption of dioxin- or PCB-contaminated breast milk are minor compared to the well-established medical benefits associated with breast-feeding (USEPA 2000b,c).
The overlapping nature of some of the effects produced in animals by some congeners of PCDDs and PCDFs makes it highly desirable to include all such "dioxin-like" compounds in risk assessments for dioxin-contaminated sites. However, the relative paucity of relevant toxicological data on non-TCDD congeners originally made such a task virtually impossible. In the early 1980s, regulatory agencies like the EPA temporarily solved this dilemma by developing the TEF approach, and recommending its adoption as a "interim science policy position for use in assessing risks associated with CDD/CDF mixtures, until more definitive scientific data are available" (EPA, 1989).
Toxicity Equivalency Factors (TEFs) are order of magnitude (factor of ten) estimates of relative toxicity of dioxin-like compounds, based on the available animal data for non-TCDD congeners and some in vitro results of limited relevance to whole animal toxicity. (The relative toxicity of 2,3,7,8-TCDD, the most toxic congener, is set equal to one.) These TEFs are often based on just one or two endpoints in a single species, usually Ah-receptor binding or enzyme induction in rodents (USEPA 1989). The Ah-receptor occurs in most vertebrate species, including humans. Binding to this receptor represents the first step in a process by which certain compounds, including PAHs and dioxins, can induce the synthesis of enzymes involved in their own metabolism. As normal physiological functions, neither Ah-binding nor the enzyme induction that may follow, are considered to be adverse health effects, per se. However, for the purposes of TEF development, they are interpreted as potential precursors of adverse effects. In the TEF approach, the measured concentrations of all 2,3,7,8-PCDDs and -PCDFs are converted to "equivalent" concentrations of TCDD by multiplying the concentration of each congener in the mixture by its TEF, thereby expressing each individual concentration in terms of hypothetical "toxicity equivalents" or TEQs. The individual TEQ values are then added together, to yield a single TEQ value for all of the PCDDs and PCDFs detected at the site (USEPA 1989; ATSDR 1998). This approach is expressed mathematically as follows:
|Dioxin Group||TEF||Furan Group||TEF|
b 1998 revised TEF
The TEFs used to calculate the TEQs were developed by the U.S. Environmental Protection Agency (EPA) and the international community in 1989 and revised in 1998 by the World Health Organization (WHO), parent organization of IARC (USEPA 1989; Van den Berg et al., 1998). Based on new studies of tumor promotion in rats and CYP1A1/A2 induction in mice, WHO raised the 1989 TEF for 1,2,3,7,8-PentaCDD from 0.5 to 1.0 and decreased the TEFs for OctaCDD and OctaCDF from 0.001 to 0.0001, respectively (Van den Berg et al., 1998). Alternative sets of TEFs do exist. Some states (California, for example) have developed their own versions, but all are strongly influenced by the International TEFs. The 1989/1998 International TEFs (I-TEFs) are listed in Table 1. (The 1998 revisions are in parentheses.)
3.2 Use of TEFs
The TEF method was originally developed as a means of arriving at approximations of the toxicity of mixtures of dioxin-like compounds that could then be incorporated into risk assessments. Risk assessments provide regulatory officials with a systematic basis for making risk management decisions and numerically ranking contaminated sites for cleanup. EPA originally considered the TEF method to be an interim approach only, one which will be replaced as soon as research yields a more accurate, practical alternative (USEPA 1989).
ATSDR uses total TEQs for screening purposes only. It is currently ATSDR’‘s policy to use TCDD-specific comparison values, expressed in TEQs, to screen all dioxin-like compounds. However, once dioxin-like compounds have been selected as "chemicals of concern" at a site, further evaluation must follow, utilizing "the best medical and toxicologic information available" (ATSDR, 1992). TEQs alone lack the information needed for the determination of probable public health implications of exposures to dioxin-like compounds. If no congener-specific data are available, no further evaluation is possible because, when health assessors know neither the identities nor the concentrations of the site-specific contaminants of concern, they are unable to avail themselves of the chemical-specific information summarized in ATSDR’‘s Toxicological Profiles for PCDDs, PCDFs, and PCBs (ATSDR, 1994, 1997, 1998).
Unfortunately, the term "toxicity equivalent" has been taken literally in many quarters and the TEF approach is now widely used by many people outside the regulatory establishment in ways for which it was never intended and is completely unsuited. EPA has indicated repeatedly that toxicity equivalency factors are not meant to be used precisely, even in a regulatory context (USEPA 1989, 1994a). It goes without saying that they cannot be used to accurately predict adverse health effects in humans.
3.3 Limitations of TEFs
The TEF approach was adopted by EPA as an interim science policy measure designed to facilitate risk management decisions (USEPA 1989). From the beginning, scientific and regulatory communities around the world cited the shortcomings in the science base supporting the TEF concept, and the latter has always been subject to revision as new experimental data became available (USEPA 1989). Some of the inconsistencies in the TEF method are summarized below.
TEFs are not precise measures of relative toxicity, and they were not designed for the prediction of adverse health effects (USEPA 1989, 1994a). EPA considers the TEF/TEQ approach to be "a useful, but uncertain, procedure." Although EPA has expressed increased confidence in TEFs and has expanded its use of the TEF/TEQ approach over the last 10 years, that agency still identifies the latter as and "interim methodology"for which "the need for research to explore alternative approaches is widely accepted" (USEPA 2000b, Chapter 9, pg 9-31). Today, TEQs based on those TEFs, though useful as screening devices, still cannot be used in the Public Health Assessment process as surrogates for actual, congener-specific data.
TEFs are order of magnitude estimates of average, relative toxicity based on highly variable data sets. TEFs are derived, by consensus, from limited experimental results that typically range over 1-to-2 orders of magnitude. These data most often come from in vitro studies and/or short-term animal studies of Ah receptor binding or P450 enzyme induction which do not represent true measures of whole-body toxicity (USEPA 1989, 1994a).
Although TEFs are defined and used as constants, the numerical value of an experimentally determined TEF will, in fact, vary significantly with a number of parameters, including: dose, duration of exposure, species, sex, strain, target tissue, and biological endpoint (DeVito and Birnbaum, 1995; Safe, 1990; USEPA 1994b). In some cases, e.g., the levels of glucose and certain blood cell types, TCDD may even have opposite effects in different species (USEPA 1994b).
TEFs can be expected to become increasingly imprecise as the conditions under which they are applied become further removed from the conditions under which they were derived. (Putzrath, 1996). TEFs are generally derived from relatively high-dose data obtained in vitro or from short-term rodent assays, but they are most often applied in the context of low-level environmental exposures in humans.
TEFs do not take into account many important biological factors. For example, the health implications of exposure to a mixture of dioxin-like compounds will depend on, among other things, the bioavailability of specific congeners and the potential for antagonistic interactions between them. TEFs, however, do not account for either (USEPA 1989, 1994a).
Finally, the assumption that all adverse effects of dioxin and dioxin-like compounds share a common mechanism, involving, as its first step, binding to the Ah receptor, is just that -- an assumption. Whether or not it is actually true, this assumption is critical to the TEF approach because it provides the theoretical basis for the simplifying inference that TEQs are additive. However, except for the chain of events leading to the induction of certain enzymes (e.g., cytochrome P-450IA1), which is not per se an adverse effect, clear evidence for such an assumption is still lacking (USEPA 1989, 1994c, 2000b, Chap 2, pg 2-19). The actual mechanism of action is still unknown for virtually all dioxin-induced adverse effects, and the available data indicate that binding to the Ah receptor is not sufficient in itself for the mediation of toxicity (USEPA 1994c; USEPA 2000b, Chap2). For example, the species most sensitive to TCDD-induced wasting syndrome (i.e., guinea pigs) is the least sensitive to CYP1A1 induction, a common measure of Ah-binding and the basis of many TEFs.
ATSDR 1992. ATSDR Public Health Assessment Guidance Manual. Agency for Toxic Substances and Disease Registry, U.S. Public Health Service, Lewis Publishers, Chelsea, Michigan.
ATSDR 1994. Toxicological profile for Chlorodibenzofurans. Agency for Toxic Substances and Disease Registry, U.S. Public Health Service, May 1994.
ATSDR 1997. Toxicological profile for Polychlorinated Biphenyls (Update). Agency for Toxic Substances and Disease Registry, U.S. Public Health Service, September 1994.
ATSDR. 1998. Toxicological Profile for Chlorinated Dibenzo-p-dioxins (Update). Agency for Toxic Substances and Disease Registry, U.S. Department of Health and Human Services. December 1998.
Aylward et al. 1996. Relative susceptibility of animals and humans to the cancer hazard posed by 2,3,7,8-tetrachlorodibenzo-p-dioxin using internal measures of dose. Environ. Sci. Technol. 30: 3534-3543.
Bertazzi, P.A., Zocchetti, C., Pesatori, A., Guercilena, S., Sanarico, M., and Radice, L.. 1989. Ten-year mortality study of the population involved in the Seveso incident in 1976." American Journal of Epidemiology 129: 1187-1200.
Bertazzi, P.A. and Domenico, Alessandro di 1994. Chemical, environmental, and health aspects of the Seveso, Italy, accident. pp. 587-632 in: Dioxins and Health (Arnold Schecter, ed.), Plenum Press, New York, NY, 1994.
Bertazzi, P.A., Zocchetti, C., Guercilena, S., et al. 1997. Dioxin exposure and cancer risk: A 15-year mortality study after the "Seveso Accident." Epidemiology 8: 646-652.
DeVito et al. 1995. Comparisons of estimated human body burdens of dioxin-like chemicals and TCDD body burdens in experimentally exposed animals. Environ. Health Perspect. 103(9):820-829.
DeVito and Birnbaum 1995. The importance of pharmacokinetics in determining the relative potency of 2,3,7,8-tetrachloro-dibenzo-p-dioxin and 2,3,7,8-tetrachlorodibenzofuran. Fundamental and Applied Toxicology 24: 145-148.
Fingerhut, M.A., Halperin, W.E., Marlow, D.A., et al. (1991). Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. New England Journal of Medicine 324: 212-218.
Gotts, R.E. 1993. Toxic Risks: Science, Regulation, and Perception. In: Chapter 12, "Dioxins and Agent Orange." Ronald E. Gotts, M.D., Ph.D., National Medical Advisory Service, Bethesda, Maryland. Lewis Publishers, 1993.
Gough, Michael (1994). Dioxin: Perceptions, Estimates, and Measures. Chap. 11, pp. 249-279 in: Phantom Risk: Scientific Inference and the Law (Kenneth R. Foster, David Bernstein, and Peter Huber, eds.), The MIT Press, Cambridge, Massachusetts, 1994.
Hayward et al. (1999). Elevated TCDD in chicken eggs and farm-raised catfish fed a diet with ball clay from a southern United States mine. Environmental Research 81: 248-256.
Huisman, M. et al. (1995). Perinatal exposure to polychlorinated biphenyls and dioxins and its effect on neonatal neurological development. Early Human Development 41: 111-127.
IARC 1997. Polychlorinated Dibenzo-para-Dioxins and Polychlorinated Dibenzofurans, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 69, International Agency for Research on Cancer, World Health Organization, Lyon, France, 1997.
Institute of Medicine (1994). Veterans and Agent Orange: Health Effects of Herbicides Used in Vietnam. Committee to Review the Health Effects in Vietnam Veterans of Exposure to Herbicides, Division of Health Promotion and Disease Prevention. Institute of Medicine. National Academy Press, Washington, D.C. 1994.
Kimbrough. 1992. Tolerable daily intake of 2,3,7,8-tetrachlorodibenzo-p-dioxin: Recent data on mechanism and animal carcinogenicity. Toxic Substances Journal 12:295-306, World Health Organization.
Koopman-Esseboom, C. et al. (1994). Effects of dioxins and polychlorinated biphenyls on thyroid hormone status of pregnant women and their infants. Pediatric Research 36: 468-73.
Koppe, J.G. et al. (1991). Breast milk, dioxins, and the possible effects on the health of newborn infants. Science of the Total Environment 106: 33-41.
Lemieux, Paul M. (1996). Emissions of polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans from the open burning of household waste in barrels. Environ. Sci. Technol., ASAP Article, Jan. 2000. Http://pubs.acs.org/cgi-bin/jtextd?esthag/asap/html/es990465t.html.
Manz, A. et al. (1991). Cancer mortality in chemical plant contaminated with dioxin. Lancet 338: 959-964.
Pluim et al. (1993). Effects of pre- and postnatal exposure to chlorinated dioxins and furans on human neonatal thyroid hormone concentrations. Environ. Health Perspect.101: 504-8.
Pluim et al., (1994a). "Dioxin and vitamin K status of the newborn", J Environ Sci Health A29(4): 793-802.
Pluim et al., (1994b). "Clinical laboratory manifestation of exposure to background levels of dioxins in the perinatal period", Acta Paediatr 83: 583-7.
Putzrath, R.M. (1996). Estimating relative potency for receptor-mediated toxicity: Re-evaluating the Toxicity Equivalence Factor (TEF) model. Regulatory Toxicology and Pharmacology 25: 68-78.
Rigo, G.H., Chandler, A.J., Lanier, W.S. (1996). The relationship between chlorine in waste streams and dioxin emissions from waste combustor stacks. ASME Research Report CRTD-Vol. 36: 1996.
Safe, S. (1990). Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFS), and related compounds: environmental and mechanistic considerations which support the development of toxic equivalency factors (TEFs). CRC Critical Reviews in Toxicology 21: 51-88.
Steenland K, et al. (1999). Cancer, heart disease, and diabetes in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. J Nat Cancer Inst. 91:779-786.
USEPA. 1989. Interim procedures for estimating risks associated with exposure to mixtures of chlorinated dibenzo--dioxins and dibenzofurans (CDDs and CDFs) and 1989 update. Risk Assessment Forum, Washington, DC. U.S. Environmental Protection Agency.
USEPA (1994a). Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and Related Compounds, Vol III, pg. 9-69,70, EPA/600/BP-92/001c, August, 1994, External Review Draft. Office of Health & Environmental Assessment, Office of Research & Development, U.S. Environmental Protection Agency, Washington, D.C., 1994.
USEPA (1994b). Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and Related Compounds, Vol II, pg. 7-139, EPA/600/BP-92/001b, June 1994, External Review Draft. Office of Health & Environmental Assessment, Office of Research & Development, U.S. Environmental Protection Agency, Washington, D.C., 1994.
USEPA (1994c). Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and Related Compounds, Vol I, pg. 3-33,36, EPA/600/BP-92/001a, June, 1994, External Review Draft. Office of Health & Environmental Assessment, Office of Research & Development, U.S. Environmental Protection Agency, Washington, D.C., 1994.
USEPA (1998). Inventory of Sources of Dioxin in the United States (EPA/600/P-98/002Aa), External Review Draft Report. United States Environmental Protection Agency, Office of Research and Development, April 1998.
USEPA (2000a). Information Sheet 1. Dioxin: Summary of the Dioxin Reassessment Science. United States Environmental Protection Agency. Office of Research and Development, Washington, DC 20460, June 12, 2000. www.epa.gov/ncea
USEPA (2000b). Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. Part II: Health Assessment for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. EPA/600/P-00/001Ae. May 2000. External Review Draft. www.epa.gov/ncea
USEPA (2000c). Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. Part III: Integrated Summary and Risk Characterization for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. EPA/600/P-00/001Ag. June 2000. External Review Draft. www.epa.gov/ncea
Van den Berg, M., Birnbaum, L., Bosveld, A.T.C., et al. (1998). Toxic Equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife. Environmental Health Perspectives 106: 775-791, December 1998.
Weisglas-Kuperus et al. (1995). Immunological Effects of Background Prenatal and postnatal exposure to dioxins and polychlorinated biphenyls in Dutch infants. Pediatric Research 38: 404-10.
Winters, Dwain (2000). Update on USEPA’‘s Dioxin Policy project. Presentation given March 17, 2000 at EPA Region IV Headquarters, Atlanta Federal Center, Atlanta, GA., by Dwain Winters, USEPA, National Program Chemicals Division (7404), Washington, DC 20460.
Zober et al. (1990). Thirty-four-year mortality follow-up of BASF employees exposed to 2,3,7,8-TCDD after the 1953 accident. Int Arch Occup Environ Health 62: 139-157.
The Agency for Toxic Substances and Disease Registry (ATSDR) issued a draft for public comment petitioned public health assessment on September 30, 1999 for the Koppers Company, Inc. (Oroville Plant) site. Between September 30, 1999, and November 29, 1999, the public had the opportunity to provide comments on the draft public health assessment. During that time, ATSDR received written comments and questions from the California Department of Health Services (CDHS), the California Department of Food and Agriculture (CDFA), Koppers Industries, and the U.S. Environmental Protection Agency (EPA). These comments/questions are presented below. Each comment is followed by a response from ATSDR.
California Department of Health Services
Comment: Our Branch, the Environmental Health Investigations Branch of the California Department of Health Services (CDHS) appreciates that ATSDR has completed the public comment document. However, we are concerned that ATSDR has mischaracterized data collected by CDHS, has not reviewed the full body of available scientific literature in assessing whether exposure levels are likely to cause adverse health effects, and a complete review of available data has not been summarized. Below we provide detailed comments of our concerns.
Response: ATSDR thanks CDHS for their comments. ATSDR has responded to the specific comments presented below and has made changes to the public health assessment as it deemed necessary in response to CDHS’‘s concerns.
Comment: Page 1: "may have resulted in reversible adverse health effects;" Page 8 "ATSDR understands that the reported symptoms have since subsided and no evidence exists that any chronic effects have occurred;" Page 13 "ATSDR does not have sufficient data...but believes that if any outcomes were triggered by exposure to domestic water supplies, the outcomes were quickly reversible. ATSDR understands that the reported symptoms have since subsided and no evidence exists that any chronic (noncancer) effects have occurred as a result of past exposures;" and Page 26 "However, ATSDR understands that the reported symptoms have since subsided and no evidence exists that any chronic (long-term) effects have occurred." These statements require a description of the date on which these statements are based and discussion of the limitations of any such evidence. No data is presented in the document to support ATSDR’‘s "belief" and "understanding." It is unclear what "evidence" was reviewed to determine that "no evidence exists." With the exception of a review of cancer statistics, which as described on page 25 is extremely limited, ATSDR has not conducted follow-up studies or, to CDHS knowledge, conducted studies of chronic effects among the residents or conducted follow up interviews with residents. If "no evidence" refers solely to a literature review and a review of exposure data, this requires clarification and greater detail and also should not be contained in the same sentence as "reported symptoms." These sentences also summarize CDHS studies of symptoms. CDHS opted to call the symptoms "acute." We consider "acute" more appropriate terminology than "reversible." "Reversible" and "subsided," terms which ATSDR uses throughout the document, implies that CDHS actually studied and observed the reported symptoms to cease. CDHS did not do this.
Response: ATSDR understands that no formal followup studies have been conducted to evaluate the current health status in the community. ATSDR’‘s reference to the apparent absence of observed longer-term health effects is based on its understanding that few, if any, complaints have been voiced by area residents in recent years. ATSDR acquired this information via communications with the EPA and CDHS throughout the public health assessment process. ATSDR revised the text to help clarify this point. It now reads "To the best of ATSDR’‘s knowledge, no recent health complaints have been received from people living/working in the vicinity of the Koppers site. No specific studies or surveys, however, have been conducted to collect health status information from area residents." (See page 14.)
In choosing the term "reversible" instead of acute or short-term, ATSDR’‘s intent was to help the public better understand that any symptoms workers/residents may have experienced in the past should not have led to long-lasting health effects. Because of limited exposure data, ATSDR could not draw any firm conclusions regarding the association between the type of effects reported (e.g., skin irritations, headaches, etc.) and possible site-related exposures. ATSDR wanted to emphasize, however, that these types of acute responses are not expected to lead to longer term (chronic) conditions-that is, the outcomes went away, or are "reversible," when the exposure ended. In this case, the term "reversible," alone, is more informative than is the term "acute", alone (CDHS’‘s preference). The term "acute" really only tells the reader how quickly the effect manifested itself after exposure; it says nothing about how long the effect might have lasted. Similarly, the term "reversible" suggests that the effect did not persist for long, but says nothing about how long it took for the effect to appear in the first place. Neither term is exclusive of the other, however, and ATSDR agrees that both terms should probably be used. Therefore, ATSDR has modified certain sections of the PHA (using both terms as necessary) to provide greater clarity.
Comment: Page 11: "Monitoring well data indicate that groundwater conditions have continued to improve since treatment began in 1993. Contaminant concentrations and the size of the plume continue to decrease." Some description of the supporting data is required here. The above two paragraphs cite a 1988 document.
Response: Throughout the health assessment process, ATSDR consulted with CDHS and EPA to ensure that we had a full understanding of the nature and extent of groundwater contamination beneath and downgradient of the site (on and off property), both presently and in the past. To evaluate groundwater conditions subsequent to 1993, ATSDR relied on data provided by EPA which included PCP data for domestic wells and monitoring wells on and off site, compiled through May/June1998. Figure 3 in Appendix A (newly added) delineates the areas where pentachlorophenol (PCP) concentrations still exceed the cleanup goal of 2.2 parts per billion.
ATSDR has added the following text to further describe post-1993 conditions:
"As can be seen in Figure 3, the "plume"of PCP contamination in groundwater has receded. Residual PCP contamination still exists to the south of the Koppers property boundary, but not near any domestic wells currently in use (Dames & Moore 1996; EPA1997a, 1997b,1998, 1999a, 1999b, 2000; HSIGeoTrans 1999)."
Comment: Page 12: "The other eight wells will be continually tested." Page 26: "Groundwater not meeting health-based cleanup goals should not be used for potable purposes." In both these instances, follow-up activities with residents should be recommended and an agency identified who will carry them out to ensure that residents are aware that water should not be used for drinking purposes.
Response: As noted in the Background section of the public health assessment (page 5), Beazer Materials and Services, Inc., (Beazer) is conducting cleanup actions under the terms of an agreement (known as a consent decree) with EPA and the conditions set forth in the "Record of Decision" (ROD) for this site. These actions include the treatment and monitoring of groundwater, both on and off property. EPA is overseeing these actions, which include the continued sampling of monitoring and domestic wells where low levels of PCP contamination are still being detected. As of December 1995, EPA determined that the area of PCP-contaminated groundwater had receded to the extent that the operation of the off-property pump and treat system could be discontinued. The residual off-property PCP contamination is being treated by use of enhanced in-situ bioremediation.
A monitoring plan is in place for on-property and off-property groundwater which generally requires quarterly sampling of area groundwater. Monitoring requirements vary depending on the concentrations of constituents detected in individual wells. If contaminants are detected at levels exceeding the ROD goals in a given well, then quarterly sampling is required. If contamination is not detected in a given well for four consecutive sampling periods (quarters), then sampling of that well will be required every 6 months. After 4 years of non-detects, a 2-year sampling of that well will be required. If at any time contamination is detected, quarterly sampling will be resumed in that well.
ATSDR has revised the text on page 12 to be more explicit about the monitoring program that is in place and overseen by EPA.
Comment: Page 13: "A review of available health effects literature revealed that virtually all studies looking at toxic effects of PCP are animal studies." There are human studies of the toxicity of pentachlorophenol (PCP) that are notable. CDHS reviewed the occupational studies and child case reports that are in the literature for the cited June 1997 CDHS Biological Monitoring and Health Interview Study (Table 16). Some of these literature citations document human death due to PCP exposure. In addition, there is a recent study of PCP workers (Hryhorczuk DO, et al. Environmental Health Perspectives106:401-408, 1998) which presents new findings and summarizes the available human studies on PCP.
Response: ATSDR agrees that this statement requires clarification. This statement was made in the context of exposures to PCP in drinking water, where ingestion is the primary exposure route of concern. Multiple human studies pertaining to PCP exposures certainly exist, but most relate to worker exposures to technical-grade PCP via inhalation or skin contact, as well as studies looking at residential exposures to PCP-treated wood. Many of these studies have no direct relevance, however, to drinking water exposures in the vicinity of the Koppers site. Few reports exist in the literature concerning adverse effects in humans ingesting PCP, the primary route of exposure.
In assessing potential public health impacts of PCP-contaminated groundwater at this site, ATSDR health assessors reviewed the literature on PCP, including the human data to which CDHS refers to in its comment. ATSDR reviewed its Toxicological Profile for Pentachlorophenol (ATSDR 1999) and also conducted a comprehensive literature search. In doing so, ATSDR sought to identify studies (human and animal) that could provide public health perspective given the information at hand regarding possible residential exposures to elevated levels of PCP in the drinking water before alternate supplies were made available.
As noted above, identified studies look primarily at worker exposures via inhalation or contact with "high" levels of technical-grade PCP, as well as studies evaluating indoor exposures in homes using PCP-treated wood. Certain factors, however, limit the usefulness of these studies: exposure and/or dose data are limited in many cases (that is, investigators do not know the PCP levels to which workers/residents were exposed or the doses responsible for the observed effects), and information on exposure to other chemicals or other risk factors are lacking (for example, it is not known whether individuals with observed health effect[s] were exposed to other chemicals or are susceptible to the observed effect[s] for other reasons [e.g., genetic susceptibility, smoking, etc.]). For these reasons, the data from these studies provided limited information on which to assess residential exposure to relatively low levels of PCP in their drinking water.
ATSDR, therefore, to a large extent, relied on data from animal studies in which the doses associated with observed adverse health effects are documented and can be used for comparison purposes in the evaluation of drinking water exposures at the Koppers site. As noted in the public health assessment, the estimated doses in area residents associated with measured levels of PCP are lower than those shown to result in adverse health effects in available studies. For added perspective, and as is described in the text (page 23), ATSDR also reviewed the limited data available on the PCP levels measured in the urine of a subgroup of area residents. These data showed that PCP levels the urine of study participants were lower than those known to be associated with health problems.
The discussion on page 13 has been modified to read as follows:
"While numerous studies exist that have identified adverse health effects associated with PCP exposures in the work setting or in homes with PCP-treated wood, little human data exist describing the effects of ingesting PCP in drinking water. Therefore, in evaluating potential health effects associated with drinking water exposures, ATSDR relied on data from experimental animal studies to provide additional perspective. Available animal studies show effects at doses above 1 milligram per kilogram of body weight per day..."
Comment: Page 14: "No firm evidence linking PCP exposure in water (drinking or skin contact) and the types of symptoms reported. Most available studies looked at exposure to PCP solutions versus diluted levels in potable water." These statements read as if ATSDR is making a distinction between drinking water and PCP solution other than the obvious difference in concentrations. To CDHS, such a distinction is not a health protective approach. There is also a case study of a hospitalized child who was exposed solely to PCP contaminated bathwater (Chapman, JB and Robson P. Lancet 1:1266-1267, 1965). Could ATSDR explain why they attribute importance to the distinction?
Response: The distinction that ATSDR was making in this statement is that it is difficult to assess the potential harm caused by diluted concentrations of PCP in water based on studies that look at observed effects associated with exposures to more highly concentrated solutions. The literature presents documentation of various adverse effects associated with these high level exposures. In some cases, effects are cited, but the concentration of the PCP solution is not provided. While these types of reports certainly tell us that harmful effects can result from dermal exposures to PCP, they do not offer the dose perspective that is key to understanding and communicating the significance of possible exposures in the vicinity of the Koppers site.
The text on page 14 has been modified to read: Studies are not available that evaluate the health effects associated with the PCP levels detected in the vicinity of the Koppers site.
The article cited by CDHS (Chapman et al. 1965) itself illustrates why ATSDR "attributes importance to the distinction" between PCP solutions and trace contamination of groundwater. The situation addressed by that case study is briefly described as follows. Thirteen days before a family moved in to it, a house and its roof timbers were sprayed with an oil-based insecticide formulation containing pentachlorophenol, beta-naphthol, and dieldrin. In the process, the cold water storage tank was contaminated, so much so that plumbers (who subsequently discovered the contamination) were called in to investigate an oily scum on the bath water. After being bathed in this water almost daily for 13 days, a little girl approximately 45 months of age was admitted to the hospital with fever, acidosis, and ketonuria. Urinary PCP levels were 6 mg/100 ml or 60,000 ppb. Nevertheless, the child recovered quickly and completely.
From the toxicological maxim, "the dose is the poison," it follows that some exposure scenarios will be more relevant than others to the site-specific conditions of exposure which are the focus of every ATSDR health assessment. The unusually high exposures described by Chapman et al (1965), for example, are not comparable to those that would result from the contact with or ingestion of groundwater containing trace amounts of PCP at the Koppers site.
Comment: Page 13-14, 1st paragraph: "inadequate evidence exists to show that PCP causes cancer in humans" and "estimated doses would be...which are 3000 times lower than the levels resulting in cancer in laboratory animals." The caveats and uncertainties of the data presented require presentation. A public health protective approach assumes that agents that cause cancer in animals may cause cancer in humans. Chemicals for which there are studies in both humans and animals demonstrate the validity of this approach. Animal studies are also limited in number and the true dose resulting in cancer in animals, and humans, is unknown. CDHS is supportive of the statement "area residents were unlikely to be exposed to site-related contaminants at high enough concentrations or for enough time to suffer long-term health effects (page 1)." However, the uncertainties in this statement are far greater and margin of safety far less than the paragraph on page 14 indicates.Response: ATSDR’‘s primary goal in its public health assessments is to put possible exposures to environmental contaminants into meaningful perspective for the public. In doing so, ATSDR strives to explain the likelihood that exposure to the detected level of a particular chemical may cause harm. ATSDR recognizes and completely agrees that a prudent public health approach is to assume that agents that cause cancer in animals may cause cancer in humans. It is also important, however, to present the weight of evidence that a given agent (in this case PCP) is associated with cancer in humans and, based on this, assess the likelihood that cancer could result from exposure to detected levels of PCP at the site. In doing so, ATSDR considers all relevant scientific data. We look at 1) the evidence of carcinogenicity in animals and humans, 2) how similarly or differently the chemical acts in humans versus animals (i.e., pharmacokinetics/dynamics, mechanism of action), 3) susceptibility and sensitivity in certain persons (e.g., children), 4) cancer registry data, and 5) any uncertainties in the available data sets.
For PCP exposures via drinking water in the vicinity of the Koppers site, ATSDR concluded that residents were unlikely to be exposed to PCP at high enough concentrations or for enough time to result in increased cancer risk. The rationale for ATSDR’‘s conclusions is as follows: First, ATSDR recognized that PCP is classified as a "probable" human carcinogen. PCP has been shown to cause cancer in laboratory animals via ingestion, but at doses significantly higher than those expected in people drinking water at the measured levels of PCP. Furthermore, it is uncertain whether PCP even causes cancer in humans. Case reports for workers exposed to PCP that show possible associations are largely inconclusive and several epidemiologic studies have shown no association (ATSDR 1999). Very little human data exist for ingestion exposures. In presenting this information, ATSDR’‘s goal is to help the public better understand the uncertainties and, again, enable them to put possible exposures into perspective.
ATSDR has revised the discussion on page 14 to better illustrate these points:
"ATSDR also looked at potential cancer threats posed by PCP at measured levels. ATSDR does not believe that PCP levels were high enough or exposure was long enough to cause cancer. PCP has been classified as a "probable human carcinogen." Although inadequate data exist to show a link between PCP and cancer in people, PCP has been shown to cause cancer in laboratory animals under experimental conditions where high doses were administered. Because possible effects resulting from lower level exposures are difficult to study, scientists are uncertain about the effects of such exposures, especially in humans. That is why screening values (e.g., ATSDR’‘s Cancer Risk Evaluation Guide) are set very low to help ensure that people are not exposed to contaminants levels even remotely close to effects levels seen in experimental studies. It is important to note that even if a 150 pound (70 kg) person were exposed to the average PCP concentration of 200 ppb detected in area groundwater (drinking 2 liters per day, every day over the course of a lifetime), estimated doses would be 0.006 mg/kg/day, which are 3,000 times lower than the lowest levels resulting in cancer in laboratory studies (18 mg/kg/day) (NTP 1989). Lastly, cancer statistics compiled for the census tracts in the vicinity of Koppers through 1989 have not shown increased cancer rates (see Health Outcome section below for further discussion on available cancer statistics, including their limitations).
As far as the "uncertainties" being "far greater" than ATSDR indicates, EPA’‘s own assessment describes the "uncertainties" associated with cancer risk estimation as follows: "the true risks are unknown and may be as low as zero." It is easy to confuse the objective of regulatory "risk" assessment with that of a "public health" information service. EPA’‘s cancer classification of PCP is based solely on "sufficient" evidence from animal studies, supported by inconsistent evidence that PCP may be weakly genotoxic. EPA characterizes the human evidence of carcinogenicity as "inadequate." Given the dose levels and durations of exposure at the Koppers site, PCP-induced cancer is not a realistic health concern.
Comment: Page 14: "Cancer statistics compiled for census tracts...have not shown increased cancer rates." As noted above, making this statement without noting the limitations of such an analysis is misleading.
Response: ATSDR agrees that discussing the limitations of the available cancer data is critical. The limitations are discussed in the discussion of Cancer Registry Data on page 26 and ATSDR has added a reference to these limitations on page 14. See also the response to the previous comment.
Unsupportable speculation has its place, but not in a public health assessment. The latter is designed to offer relevant answers to specific questions, using the best medical and toxicological information available (ATSDR 1992c).
Comment: Page 18: "In looking at the data collected in 1987 and 1988, ATSDR found that approximately 98% of the TEQ in eggs consists of the hepta-, hexa-, and penta-forms of dioxins/furans. All known effects, ..." CDHS considers ATSDR’‘s approach of examining the toxicity of the individual congeners unnecessary. The described TEQ approach is an internationally accepted approach for equating the other forms of dioxins/furans to TCDD and assumes that the other forms have much lower toxicity.
Response: ATSDR recognizes that the TEQ approach is an acceptable and convenient screening approach for assessing potential dioxin/furan hazards. However, as stated on page 18 and in Appendix E (page E-5), because of uncertainties regarding differences in congener toxicity and the mechanisms of toxicity in humans, TEQs cannot be used as a tool for predicting adverse health effects in humans. When screening values are exceeded, further evaluation is needed. Such an evaluation must look at all relevant toxicologic and medical information. In this case, that means 1) congener-specific exposure data, 2) congener-specific effects data, and 3) established principles of toxicology and medicine. That is why, in its assessment of dioxins in home-raised eggs, ATSDR also reviewed the congener-specific data. The purpose was to describe the uncertainties to the public and to provide additional perspective regarding possible dioxin exposures.
Perhaps some quotes from an EPA assessment of TEFs may help clarify the scientific validity of equating forms of dioxin/furan to TCDD and predicting the health effects of those mixtures in humans. The following quotes are all from EPA (1989), "Interim procedures for estimating the risks associated with exposures to mixtures of chlorinated dibenzo-p-dioxins and -dibenzofurans (CDDs and CDFs) and 1989 update," EPA/625/3-89/016.
"...many of the short-term results seen in murine systems are not observed in rat systems. Also, the connection between the enzyme induction response, which supports several of the TEF values, and several of the toxic endpoints manifested by CDDs/CDFs, is unclear. Other mechanisms of action, e.g., effects on vitamin A synthesis and estrogen-like activity, have been suggested as playing an important role in the toxicity of CDDs/CDFs. These continuing elements of uncertainty in the TEF approach highlight the need to treat the approach as ‘‘interim,’‘ that is, one that needs to be further buttressed by experimental data and eventually replaced with a more direct biological assay." (Page 8)
"...these estimates [i.e., TEFs for Hp- and Octa- CDDs/Fs] ignore the issue of relative bioavailability of the CDD/CDF congeners." (Page 10)
"With the exception of 2,3,7,8-TCDD, the 2,3,7,8-HxCDDs, and 2,3,7,8-TCDF, the TEFs are not based on the results of major animal (reproductive, carcinogenic) studies. Generally, TEFs are based on estimates of the relative toxicity in in vitro tests whose relationship to the chronic effects of concern is largely presumptive." (Page 13)
"...research holds the promise of removing the need for any TEF scheme. This is particularly important in the light of the emerging data showing that some of the CDDs/CDFs and related compounds can exhibit antagonistic effects (Safe, 1987), a possibility that is ignored in current TEF approaches."
Although the use of the TEF approach has expanded considerably since 1989, the scientific limitations of the TEF method remain essentially the same more than 10 years later. The concept of "toxicity equivalence" was not developed for the realistic prediction of human health risks from chemical exposures and should not be interpreted literally; it was identified by EPA in its 1994 risk assessment of dioxin as one of its seven "key assumptions and inferences" (HAD 1994, Vol III, pg. 9-73).
ATSDR believes that an examination of individual congeners is of critical importance. Without it, "further evaluation" (which, according to ATSDR guidance, must follow whenever maximum, site-specific concentrations of a contaminant exceed the relevant comparison values) would be impossible. Only by taking into account all of the relevant chemical, biological and pharmacologic factors that are not accounted for in the TEF approach is it possible to produce a health assessment that goes beyond regulatory assumptions in an effort to ascertain (as nearly as possible) the realistic exposures and health risks in people exposed under site-specific conditions.
Comment: Page 18: "Humans are thought to be 10 times less dioxin-sensitive than most laboratory animals," and Page 24: "Furthermore, humans are thought to be at least 10 to 100 times less sensitive to dioxins than are rodents." These statements are in disagreement with a recent review article which states "humans and animals have similar degrees of sensitivity." (Grassman JA, et al. Environmental Health Perspectives: 106(2):761-775, 1998).
The main reason for the appearance of a disagreement is that the first statement that "humans are thought to be 10 times less dioxin-sensitive" was a quantitative conclusion based on toxicological principles and research while the comment attributed to Grassman et al. was a qualitative statement based largely on regulatory assumptions and methodologies. Grassman et al. do assert that humans respond "similarly" and have "similar" degrees of sensitivity to dioxin-induced effects. However, those assertions were not actually supported in the body of that article. In the review article by Grassman et al., the authors briefly describe a variety of real and presumed effects (adverse and otherwise) of dioxin exposure in animals and humans, but make no real attempt to either establish causality in humans or directly compare the doses that produce the same effect in humans and animals. Nor do the authors ever clarify their use of the term "similar." Such clarification is available, however, from the reference (# 17) cited by Grassman et al. as the source of their statement regarding the alleged similarities between human and animal responses to dioxin and dioxin-like compounds. In that cited article, DeVito, Birnbaum, Farland, et al. (1995) stated that, as they used the term, the word "similar" meant "within a factor of ten."
Unlike Grassman et al.(1998), DeVito, Birnbaum, Farland, et al. (1995) did compare effective doses in an effort to support their conclusions. However, those efforts depended heavily on a number of assumptions and practices that are not generally employed by toxicologists outside of the regulatory community. These include:: (1) the definition of "similar" as "within a factor of ten," (2) a focus on non-adverse effects, (3) the use of "body burden" (i.e., the concentration in fat) as the toxicologically-relevant measure of dioxin dose, (4) a literal interpretation of TEQs as toxicity "equivalents," and (5) certain assumptions about the essentially unknown mechanism of action of dioxin, assumptions which are crucial to the TEF approach to risk assessment but remain largely unproven. Thus, the conclusions of DeVito et al. were speculative in nature. This did not constitute a problem with the DeVito et al. paper, because the authors themselves acknowledged many of the limitations of the assumptions on which their conclusions were based. More importantly, those conclusions were not presented as scientific findings that clearly established either the existence of or the mechanism behind dioxin-related adverse health effects in humans. Those authors’‘ primary purpose was to present plausible, theoretical support for the validity of the TEF concept, for regulatory purposes. This they accomplished. In so doing, however, they openly acknowledged that "chloracne is the only toxic effect induced by dioxins for which there is unequivocal evidence linking exposure to effect in humans" (DeVito, Birnbaum, Farland, et al. 1995, middle of pg 1). (Additional clarification of this and other dioxin-related issues can be found in an Appendix E.)
Comment: Page 18, next to last paragraph, and Page 24, last two paragraphs: These paragraphs discuss the toxicity of dioxins and furans and the concentrations found in eggs. These paragraphs should include the recent human cancer studies and the International Agency for Research on Cancer evaluation of TCDD as a human carcinogen. The cited CDHS reference (Goldman LR, et al., Environmental Health Perspectives: 108:1-7, 2000) compares levels observed in one of these studies to those found among the egg and beef consumers. These cancer studies estimated blood levels among a group with a documented elevated rate of cancer to be approximately equal to those found among the egg and beef consumers. Also pertinent is that researchers have estimated that "some individuals may respond to dioxin exposures with cancer and noncancer effects at body burdens within one to two orders of magnitude of those in the general population" Devito MJ et al., Environmental Health Perspectives: 103(9):820-831, 1995).
Response: Appendix E containing general information on dioxin and dioxin-like compounds has been inserted into the final version of ATSDR’‘s Koppers Public Health Assessment. It addresses, among other things, the issues of 1) human cancer studies, 2) IARC’‘s re-classification of 2,3,7,8-TCDD as a human carcinogen, and 3) the limitations of using "body burden" as a dose metric to establish toxicological relationships. In Appendix E, ATSDR endeavors to make clear the distinction between established, objectively-demonstrable, toxicologic and human health observations, on the one hand, and prudent, politically acceptable, public health policy on the other. Nowhere is this distinction more important than in discussions of dose-response relationships. The reader is referred to Appendix E for details. The rest of the following response specifically addresses each of the 3 separate parts of the above comment concerning 1) IARC’‘s re-classification of TCDD, 2) The CDHS reference, Goldman et al. (2000), and 3) the body burden statement from DeVito (1995).
The epidemiological evidence to date remains inadequate to establish a cause-and-effect relationship between TCDD exposure and cancer in humans. Both IARC and EPA continue to regard the epidemiological evidence for TCDD-induced cancer in humans as "limited." Thus, IARC’‘s re-classification (like EPA’‘s proposed re-classification) is based on "limited" evidence in humans, and "sufficient" evidence in animals. Historically, these two criteria have been used to justify designating substances as "probable" human carcinogens. In the case of TCDD, the classification as a "known" human carcinogen depended on additional considerations, which IARC referred to as "supporting evidence." These considerations included 1) evidence interpreted as implying the existence of a common, Ah receptor-dependent mechanism of carcinogenic action in both humans and animals, and 2) the observation of "similar" tissue concentrations of TCDD in animals and human populations exhibiting elevated rates of cancer. (The actual blood levels compared, in this case, ranged from 301 to 32,000 in 4 epidemiological studies and from 1,500 to 10,000 ppt in rats (IARC, 1997, pg342-3).) The inferences drawn from both lines of this "supporting evidence" depended heavily on assumptions central to the TEF approach to the assessment of risk from exposure to dioxin-like compounds. These assumptions and inferences, while useful for the regulatory purposes of screening and risk assessment, lack both the precision and the necessary science base to make them valid tools for predicting effects across species. (See response to previous question.)
For example, the regulatory assumption that most, if not all, of dioxin’‘s adverse effects are mediated by binding to the Ah receptor, the actual mechanism of virtually all of those adverse effects remains essentially unknown. Also, while concentrations in body fat or blood lipids represent excellent biomarkers of exposure to TCDD, they do not constitute reliable biomarkers of effect. (Grassman et al., 1998, which was cited in the previous comment, specifically warns that "cross-species comparisons of tissue concentrations should be approached with caution.") Notwithstanding the regulatory assumption that dioxin concentrations in adipose tissue (used as a surrogate for "body burden") are directly relevant to adverse health effects, there is no evidence that those effects are actually caused by the dioxin that has been sequestered in fat. On the other hand, there is substantial evidence that sequestration in fat actually protects the organism from dioxin-related adverse effects by drawing dioxin away from target tissues like the liver (Kimbrough 1992). There is even a good linear correlation between the LD50 of dioxin in many different species and the log percentage total body fat in those species (Geyer et al. 1990, Toxicology 65:97-107). Based on this relationship, the estimated 30-day LD50 in humans would be about 6,000,000,000 pg/kg. Children exposed during the 1976 accident at Seveso, Italy (Zone A) had the highest blood levels of TCDD ever recorded (up to 56,000 ppt compared to a U.S. average at the time of perhaps 7 ppt) and, at Zone A, were exposed to an estimated mean of 3,125,000 pg TCDD/kg. Nevertheless, after more than 15 years of medical follow-up, chloracne (which resolved in most cases within a few years) remains the only adverse health effect suffered by these individuals that can be attributed to their TCDD exposure. In the context of such information, speculations based on body burden comparisons between animals and humans (or between humans with and without elevated cancer/mortality rates) would have no impact on ATSDR’‘s assessment of the potential for dioxin-related health effects in individuals exposed at the Koppers site.
The limitations of the "body burden" approach to dioxin risk assessment also apply to the quoted comment from DeVito et al.(1995). The statement to the effect that TEQ body burdens within one to two orders of magnitude of those in the general population may be causing cancer and noncancer effects in some people is essentially speculative. Such speculations are commonly necessary to support risk assessments and to justify regulatory decisions. However, they are not particularly relevant to the assessment of the public health effects that are likely to occur in people potentially exposed to site-related contaminants under site-specific conditions.
Finally, it should be noted that Goldman et al. did not actually claim that estimated blood levels among a group with a documented elevated rate of cancer were "approximately equal" to those found among the egg and beef consumers. They observed, rather, that the difference between the means of the high exposure and comparison groups (48 pg ITEQ/g) in their own study was "approximately equal" to the difference between background and the lowest exposed group (40 pg ITEQ/g) with significantly elevated cancer rates in the Hamburg occupational mortality study by (Dieter) Flesch-Janys et al. (1995). (Note that the first author’‘s last name was left off in the cited reference in Goldman et al, 2000.) In the latter case, quantitative estimates of PCDD/F exposure in the whole cohort and specific exposure groups were based on extrapolation and back-calculation from levels measured in blood lipids and adipose tissue in a subgroup of workers (N=190) several years after exposure, and the referents were gas workers. In the Goldman et al. study, the mean exposure in the high exposure group (i.e., egg and beef consumers) was derived from blood samples from only 5 individuals, and PCDD/F levels in the rural comparison group were generally lower than levels in the urban U.S. population. Thus, neither the exposure nor the comparison groups used in these two very different studies are comparable to one another. Also, regarding the implication of significantly elevated cancer rates in the lowest exposure group in the Hamburg occupational mortality study, it should be noted that the relative risks for all cancers combined - the authors did not look at specific types of cancer - remained non-significant or borderline significant until the fifth quintile (highest exposure group) which included estimated exposures ranging from 545 to 4,362 ppt TEQ (Flesch-Janys et al., 1995). In any case, for all of the reasons mentioned previously, group average concentrations of dioxins in fat are no more predictive of adverse health effects than are individual concentrations in fat.
As ATSDR has stated elsewhere in this Public Health Assessment, exposures to products potentially contaminated with elevated levels of dioxins should be limited wherever reasonably possible. (See pages 2, 19, 26, and 27). However, that recommendation is predicated on prudent public health policy and not on any scientific data or speculative estimates suggesting that adverse health effects are likely to occur now or in the future at the levels of exposure documented at Koppers/Oroville. Such distinctions are a critical component of effective risk communication and, hence, of ATSDR’‘s Public Health Assessments.
Comment: Page 24: "Blood levels in the highest consumers group (beef and egg) were less than 9% higher than the high end of this range." This comparison is unnecessary and should be deleted. The cited CDHS article (Goldman LR, et al., Environmental Health Perspectives: 108:1-7, 2000) compares the beef and egg consumers, not only to appropriate age and gender matched rural comparison subjects but also to an urban population, with an approximately equal age and gender distribution. It is unclear whether the ATSDR cited reference groups are age and gender appropriate. Further, means should be compared; not the "high end" as the ATSDR cited reference groups may be much larger and more likely to have outliers than the smaller egg and beef study group.
Response: The primary function of a public health assessment is to effectively communicate to potentially exposed residents a meaningful and useful perspective on the probable health implications of their site-specific exposures. The point of ATSDR’‘s comparison was that the highest blood levels detected at Koppers were not very high, relative to the range of background exposures in the United States and were still quite low compared to levels known to produce effects in human beings. And, a congener-specific analysis demonstrated that the more highly-chlorinated (and, therefore, less toxic) congeners predominated. ATSDR concurs with CDHS’‘s cautionary recommendation that exposures be limited wherever possible. However, for effective risk communication, it must also be acknowledged that the documented levels of exposure in Oroville residents would not be expected to result in adverse effects and, therefore, need not occasion alarm. Any such alarm would be much more likely to lead to detectable adverse health effects (e.g., stress-related, psychogenic illness) than would any of the site-specific dioxin exposures. ATSDR considers that the cited statement provided the appropriate perspective.
Comment: Page 24: "To the best of ATSDR’‘s knowledge, however, no adverse health effects have been reported in the study group, to date." This statement refers to a very small group of residents from whom biological samples were collected. This statement should be deleted as it implies that the study group has been actively followed over time. The study group has not been followed up by any government agency, including ATSDR. Follow up of this small study group would be inappropriate as the study was designed to study exposures, not health effects. Although the biological levels were elevated and suggest an elevated cancer risk, the study group is too small to expect to be able to observe that elevated cancer risk.
Response: ATSDR agrees that it should be clearly stated that the group has not been followed up medically and that the small study group would limit the observation of elevated cancer rates. Such a statement has been added to this discussion.
California Department of Food and Agriculture
Comments related to the Petitioned Public Health Assessment for Koppers Company, Incorporated (Oroville Plant) are limited to two areas of concern, namely methodology and recommended action:
Comment: The report has successfully related risk estimates to human health quantitatively, however it acknowledges that duration of exposure has not been incorporated into the exposure assessment. However important exposure duration is for humans, it is also important to interpret International Toxicity Equivalents (ITEQ) in light of the fact that the average life span of backyard poultry may approach a duration of up to ten years and may be of greater risk.
Response: ATSDR was not attempting to relate risk estimates to human health quantitatively. Apart from being convenient as screening values, "risk" estimates generated using regulatory methodology have little or no relevance to the process of human health assessment. They do not necessarily give a realistic prediction of the true risk, which is unknown and may be as low as zero (EPA 1986). In addition, they are often confused by the public with estimates of actuarial risk which, unlike quantitative cancer risk estimates, are based on actual incidence data in humans. TEQs, too, are of limited relevance to the public health assessment, apart from their use in preliminary screening values. The TEF method does not consider many of those pharmacologic factors (e.g., bioavailability, potential for antagonistic interactions, and congener-specific metabolism, half-lives, and effects) that are of paramount importance in any scientifically valid toxicologic evaluation of the probable health implications of site-specific exposures. Also, ATSDR advising against the sale and continued consumption of home-grown eggs at the two index homes would effectively avert the "greater risk" referred to in this comment.
Comment: Normal (control) background levels of dioxin congeners for Butte County have not been presented and so it is difficult to assess the scientific significance of the report findings in light of this important basis of comparison.
Response: Should such background data become available, ATSDR will include it in an addendum to the PHA. However, such data are not pertinent to the central question of whether or not the identified site-specific exposures are likely to have an adverse impact on public health. ATSDR considers that the available information was sufficient to support a reasonable answer to that question, within the limits of current knowledge and technology.
Comment: A site map with the locations of the two index backyard sites and the remaining 23 sites that were sampled would be very useful to correlate with the demographic, water, and soil maps provided.
Response: ATSDR has not been able to identify a map that clearly depicts these locations. However, ATSDR has created a map in Appendix A using a 1989 CDHS reference (Foraging Farm Animals as Biomarkers for Dioxin Contamination) for contaminated egg locations; although the map in this reference is not clear, a general idea of egg sampling locations is can be determined.
Comment: The report states that exposure assessment considered the following routes of human uptake: Ingestion, Inhalation, and Dermal. In addition, the report states that "chemicals detected at comparison values simply require a more detailed evaluation of site-specific exposure conditions." If such a site-specific assessment was done, it would be important to correlate such results with the levels detected in a geographic origin of animals, animal products and animal by-products noted in the report in the interest of selecting specific risk management strategies appropriate at each site.
Response: This comment references ATSDR’‘s assessment methodology, which outlines in general terms how ATSDR evaluates environmental and health outcome data. Section B of the Discussion section presents the findings of the site-specific analysis which focuses on the following five potential exposure pathways (as outlined on page 8):
- Ingestion and skin contact with PCP.
- Exposure to smoke during past fire episodes.
- Ingestion of homegrown farm products.
- Contact with on-site soils.
- Contact with surface water and sediment in streams along the site perimeter.
Site-specific exposure information, including exposure assumptions, are included in these discussions (e.g, drinking water assumptions and egg consumption patterns for area residents). It should be emphasized, however, that the primary purpose of a public health assessment is to identify site-specific exposures and determine their probable implications for the health of local residents. The ultimate origin of the substances to which those residents are exposed is of secondary interest only. Such knowledge is not necessary for the determination of probable health effects, but it could be very useful in the effort to determine the most appropriate measures for limiting exposures.
Comment: No mention is made in the report of sampling insects which are a staple food of backyard poultry.
Response: To ATSDR’‘s knowledge, no efforts were made to sample insects, as the latter were not considered to be potential sources of significant human exposure to PCP or dioxin. It is expected that poultry at Oroville/Koppers would most likely be exposed to these contaminants while pecking at contaminated soils. The detected levels in off-site soil, though only modestly elevated, would have been bio-accumulated in poultry foraging on such soils.
Comment: Human disease rates are not adjusted for age, sex, or race.
Response: ATSDR recognized this limitation in interpreting the available cancer registry data. The potential limitations associated with not adjusting the cancer rates for age, sex, and race are briefly described in the public health assessment on page 26.
Comment: Sample size for animal products was not defined and as a consequence the statistical significance of the report findings is not apparent.
Response: The purpose of the public health assessment is to put past, current, and potential future exposures to contaminants potentially related to the Koppers site into perspective for the public. It is a largely qualitative exercise based on data collected from various sources. ATSDR agrees that the representativeness of the available sampling data is an important factor for interpreting results. (Table 5 summarizes the animal product data.) However, it should be emphasized that ATSDR typically bases its health calls on the maximum concentrations detected and worst-case exposure scenarios, both of which are intentionally unrepresentative.
Comment: Education of backyard producers in the use of proper management of livestock and poultry is a major emphasis of current and future risk mitigation recommendations. Government agencies with expertise in poultry management and diseases such as the CDFA are excellent sources of such assistance.
Response: ATSDR acknowledges the assistance that CDFA and other agencies can offer. ATSDR has learned that CDFA is currently participating in a Task Force, led by the CDHS, to develop specific guidance on the proper management of backyard livestock and poultry and to develop a comprehensive educational program to reach all potentially affected populations. Therefore, ATSDR has modified the recommendations/public health action plan in the public health assessment to describe these activities.
Comment: Recommendation #5: "No untested poultry products from these backyard farms should be sold." The report is unclear as to whether or not sales of poultry products from backyard flocks are occurring. If so, it does not explain what mechanisms are in place to prevent untested poultry products from affected backyard poultry from being sold.
Response: The advisories issued by CDHS have advised backyard producers not to sell or consume backyard products. As stated on page 19 of the public health assessment, in light of the uncertainties pertaining to dioxin toxicity and exposure levels, it is prudent public health practice to limit exposures to the extent possible and ATSDR, therefore, recommends that residents continue to adhere to the CDHS advisory. ATSDR recognizes that no mechanism is in place to strictly regulate or require testing of these products and encourages the type of educational programs described in the previous response.
ATSDR’‘s evaluation of possible exposures to the levels of dioxins/furans detected in backyard poultry in the vicinity of the Koppers site presented in the public health assessment attempts to provide readers with some perspective on the likelihood of adverse health effects and what is and is not known about low-level exposures to dioxins/furans.
Comment: Page 1, Paragraph 3: It is unclear if "contaminants detected on-site..." refers to all or selected modalities. With respect to breathing zone air, sampling data from industrial hygiene air monitoring studies for the site would not support the presence of benzene as an airborne contaminant."
Response: This paragraph summarizes what was found in on-site groundwater, soil and/or air. Text discussions and accompanying tables clearly indicate which contaminants were detected in which medium. Benzene was detected in on-site groundwater only, based on the sampling data that ATSDR reviewed (see Table 2).
Comment: Page 14, On-site Air: Industrial hygiene air monitoring data collected on a cross section of plant employees during routine operations/activities revealed no elevated levels of BTS/Naph, CTPVs, Arsenic, Copper, Chromium, and PCP.
Response: ATSDR does not typically review industrial hygiene data because the protection of workers falls under the jurisdiction of the Occupational Safety and Health Administration (OSHA). If air data are available in the vicinity of the aeration lagoon, the two wastewater spray field, and spill areas, these data would be helpful in helping to assess possible past exposures to air on and in the vicinity of the site. However, the greatest community concern related to air exposures was possible health effects resulting from releases during fire episodes, for which no air data are reportedly available.
Comment: Page 15, Paragraph 2: Extensive wipe sampling and limited air sampling was conducted on-site immediately following the fire and during associated cleanup activities.
Response: Wipe sampling would provide some additional information regarding residual levels of contaminants following the fire, but would not fill the primary data gap-that is, levels of potential contaminants in air to which area residents and workers may have been exposed at the time of the fires.
Comment: In general, the report appears to be factual and technically sound. However, one area that is not clearly described involves the relationship of dioxins/furans in backyard-raised eggs and livestock to the subject plant site. A transport pathway for these compounds from the Oroville Plant to the subject receptors has not been established. In fact, data summarized on page 16 of the report (from CDHS 1997, 1999; Goldman et al. 1999) indicates that dioxins were detected in samples collected both before and after the fire incident at the site, which was presumed to be the source of these compounds at the time samples were collected. This suggests that there may not be a relationship between the presence of dioxins/furans in eggs and livestock and activities at the Oroville Plant, and that there may be another source of these compounds that is unrelated to the Oroville site.
Response: ATSDR recognizes the fact that pre and post-fire levels in a single cow sample were at similar levels, suggesting the existence of a pre-fire source of dioxins. ATSDR also recognizes that home-produced eggs and poultry revealed elevated levels of dioxins/furans despite the presence of only trace levels of dioxins/furans in areas soils. The latter was not entirely unexpected, however, considering the inevitable bioconcentration of dioxins in chickens pecking at contaminated soils. Note, also, that ATSDR’‘s primary role in this assessment was to provide a public health evaluation of the reported levels of dioxins/furans in home-produced eggs/livestock in response to community concerns--not to perform an extensive evaluation of the potential dioxin/furan source(s).
U.S. Environmental Protection Agency, Region IX
Comment: General - EPA uses the terminology on-property and off-property vs. on-site and off-site (under CERCLA the site is wherever contaminants are found/migrated to).
Response: ATSDR generally uses "on-site" and "off-site" to designate contaminant location. On-site and off-site designations have been changed to "on-property" and "off-property" as appropriate in the PHA, primarily in groundwater discussions.
Comment: Page 5, third paragraph - As of December 1995, EPA determined that the area of PCP-contaminated groundwater had receded to the extent that the off-property Pump and Treat system operation could be suspended. The residual off-property PCP contamination is presently beginning treated by use of enhanced in-situ bioremediation. (Continued use of P&T off-property would have required the installation of new extraction and re-injection wells piped to the P&T to treat the residual plume. Relocating the off-property P&T remains a remote option if other remediation efforts are not successful.)
Response: In-situ bioremediation has been referenced in the PHA as a result of this comment.
Comment: Page 5, fourth paragraph - Change 100,000 to 107,000 cubic yards.
Response: 100,000 has been changed to 107,000 cubic yards as a result of this comment.
Comment: Page 12, first paragraph - Modify last sentence. The remaining 8 drinking water well owners with PCP contamination will continue to be paid an OWID allotment until it is demonstrated that their wells meet the ROD PCP criteria for 12 consecutive months. (Sampling of these eight wells is not performed continually. They will be sampled prior to ending the OWID subsidy. Five of these deactivated drinking water wells are in the area of the residual plume.)
Response: This sentence on page 12 has been modified in the PHA.