TL: ACHIEVING ZERO DIOXIN: An emergency strategy for dioxin elimination SO: Michelle Allsopp with contributions from Ruth Stringer, Joe Thornton, and Pat Costner, Greenpeace International, (GP) DT: September, 1994 1. Contents 2. Summary 4 Risk Assessment 4 Dioxin~a by-product of chlorine chemistry 5 International Regulations 6 An emergency strategy 7 3. Dioxins in the environment and in humans 9 Chemical characteristics of dioxins 9 Levels of dioxin in the environment 9 Effects on fish and wildlife 11 Human exposure 11 Food 11 Soil and sediments 12 Levels in human blood, tissue and breast milk 13 4. Toxicity of dioxin 14 How dioxin exerts toxic effects 14 Effects on reproduction and development and the immune system 14 Developmental toxicity 15 Effects of accidental/occupational exposure 15 Effects on development occur at very low doses 16 Studies of effects on humans at current background levels 16 Reproductive Toxicity 17 Effects on the immune system 18 Carcinogenicity 18 5. Risk Assessment 20 WHO and Health and Welfare Canada method of risk assessment 20 US EPA method of risk assessment 21 WHO and Canadian assumptions 21 1. Dioxin thresholds? 22 2. Most sensitive end-points in risk assessment 24 3. Body burden 24 Future Risk Assessment 25 6. Formation and sources: dioxin and chlorine chemistry 26 Combustion sources 26 Incinerators 26 Vehicle fuels 28 "Natural" combustion sources 29 Industrial sources 30 Processes in the pulp and paper industry 30 Processes in the chemical industry 30 Chlorinated aromatic chemicals 31 Synthesis of aliphatic compounds: PVC 31 PVC production 31 PVC incineration 34 Recycling of PVC-containing products 34 PVC in fires 34 Synthesis of aliphatic compounds: chlorinated solvents 34 Inorganic chlorochemical processes 35 Processes involving chlorinated intermediates: pesticides 36 Metallurgical processes 37 7. Phasing out dioxins 38 Immediate priorities: major dioxin sources 39 1. Incineration and other combustion sources 40 2. Pulp and paper 40 3. PVC 40 4. Chlorinated aromatic chemicals 40 Secondary actions 40 1. Chlorinated solvents 40 2. Chlorine-related pesticides 40 3. Metallurgy 40 4. Water and wastewater treatment 41 5. Remediation 41 Economic implications 41 8. References 42 ********************************************************** 2. Summary Dioxins are a class of chemical compounds which are extremely toxic to animals and humans, and have been characterised as being some of the most potent "man made" toxic chemicals ever studied. Consequently, in recent years, there has been much public concern and intense scientific interest about these chemicals. During the week of September 12th 1994 the US Environmental Protection Agency (EPA) intends to release its reassessment of environmental and human health risks from exposure to dioxin. It is expected to conclude that exposure to dioxin poses a largescale, long term threat to public health and the environment~not only due to cancer risks but because of possible birth defects and damage to immune systems. In the light of these findings Greenpeace has published this report~outlining how and why dioxin is affecting the environment and human health. It identifies the key sources from industrial uses of chlorine and highlights dioxin sources not adequately considered by the EPA. Finally this report suggests strategies for eliminating all these sources. However, it is thought that the EPA reassessment will widen and deepen the debate about exposure levels to dioxin and seriously challenge current assumptions. Risk Assessment In December 1990, the World Health Organisation (WHO) produced a document on human health and dioxin. In this document, a ~tolerable daily intake~ for dioxin of 10pg/kg body weight/day (pg/kg/d) was proposed. Tolerable daily intake (TDI) is used to define the maximum dose levels for lifelong exposure that may be regarded as not detrimental to human health. The TDI proposed by WHO has been adopted by many European governments. The Canadian government has also legislated a TDI of 10pg/kg/d based on research by Health and Welfare Canada. However, using a different method of risk assessment for dioxin, the US EPA had set a value for the ~acceptable daily intake~ (ADI) of dioxin of 6 fg/kg/d. This is 1670 times lower than the WHO and Canadian TDI. Recent studies carried out under the auspices of US EPA reassessment of dioxin, indicate that assumptions made in the calculation of the WHO and Canadian TDI are not consistent with current research findings. This means that the TDI of 10pg/kg/d may be too high, and therefore may not be protective of public health. The main controversy surrounds the assumption made in the derivation of WHO and Canadian TDI, that a threshold or "safe dose" of dioxin exists, below which no adverse effects on health will occur. Recent studies, however, do not support the idea that such a threshold exists for dioxin. Also, health effects caused by dioxin are now known to occur at lower doses than those that were assumed in the calculation of the TDI. Consequently, it is very probable that the TDI set by WHO and Health and Welfare Canada is incorrect and therefore is not protective of human health. However, no reassessment of the WHO TDI is planned for the future. In 1985 and 1988 the United States Environmental Protection Agency (US EPA) prepared assessments on human health risks from exposure to dioxin. However, in April 1991 US EPA Administrator William Reilly announced that another reassessment of dioxin would be undertaken. The reasons for this reassessment were two-fold: Firstly, a National Institute of Occupational Safety and Health (NIOSH) cancer mortality study in US workers, published by Fingerhut et al.(1991), showed that 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), the most toxic dioxin chemical known, appeared to cause cancer in humans. Secondly, the Banbury Conference on dioxin toxicity held in October 1990 in Cold Spring Harbour, New York, produced evidence of several other important findings on dioxin toxicity, including the fact that humans and experimental animals respond to dioxin similarly, and effects in humans could be anticipated by certain effects in animals. In view of these two events, US EPA publically announced plans for the reassessment of dioxin on November 15, 1991 (Federal Register 1991). Aims of the reassessment included the development of a new model for estimating human health risk from dioxin, updating research on health assessment and exposure to dioxin and supporting research to characterise ecological risks in aquatic ecosystems. Dioxin~a by-product of chlorine chemistry Dioxins are produced as unintentional by-products from the many processes in which chlorine and chlorine-derived chemicals are produced, used and disposed of. Industrial emissions of dioxins into the environment can be transported long distances on atmospheric currents and to a lesser extent in rivers and oceanic currents. Consequently, dioxins are now globally ubiquitous. It is estimated that even if production ceases, levels in the environment will take years to decrease. This is because dioxins are persistent, taking decades or centuries to degrade, and can undergo continual recycling throughout the environment. Human exposure to dioxins is almost exclusively from food intake, especially from meat, fish and dairy products. Unusually high exposure to dioxins in humans following for example accidental/occupational exposure, together with experiments in laboratory animals, have shown effects of dioxin on health include developmental and reproductive toxicity, effects on the immune system and carcinogenicity. Even more disturbing are findings from recent studies which show that concentrations of dioxins in human tissue in the general population (of industrialised countries) are already at "or near" those levels where health effects may occur. Recent research on health effects from dioxin indicates the following important points: 1. In fish, birds, mammals and humans, evidence shows that the developing foetus/embryo appears to be very sensitive to toxic effects of dioxin. Developmental effects in humans, seen after high accidental/occupational exposure to dioxins include: pre-natal mortality; decreased growth; organ dysfunction, for example involving effects to the central nervous system such as impairment of intellectual development; functional alterations including effects to the male reproductive system. 2. Animal and/or human studies have shown that some effects, for example cellular changes in the immune system, changes in the levels of male sex hormone testosterone, and changes in other enzymes and hormones, may be occurring in humans at, or near to, current levels (body burdens) of dioxins found in the general population of industrialised countries. Such effects could lead to adverse effects on human health. Members of the population who have higher than average exposure to dioxin, for example through having a high fish or sea mammal diet, are more at risk from such adverse effects including the possibility of reduced sperm count, impairment of the immune system and endometriosis in women. 3. Biological effects from dioxins appear to depend on the concentration present in a target organ over a critical time period rather than on dose. Animal experiments have shown that exposure to very low doses of dioxin during an extremely short critical period during gestation is sufficient to cause detrimental health effects on the foetus. 4. In industrialised countries, levels of dioxins in breast milk often result in nursing infants having dioxin intakes far in excess of the TDI proposed by WHO. This becomes of even greater concern when it is considered that health risk assessments of dioxins do not take other chemicals into account such as polychlorinated biphenyls (PCBs) which humans are exposed to. The effects these chemicals have on given health points may be additive to dioxin or synergistic, i.e. produce a larger than additive enhanced effect. 5. Evidence from studies of occupational/accidental exposure to dioxin in humans together with animal studies indicatesthat dioxin causes cancer in humans. US EPA estimated that current background exposure to the general population results in cancer risks ranging from 1 in 1000 to 1 in 10,000. International Regulations The phase out of persistent, toxic and bioaccumulative pollutants in the environment has already been addressed at several international conventions including: the Third International Conference on the North Sea (1990) in which it was agreed that dioxin and other named substances should be reduced of the order of 70% or more; the Paris Convention (1992) in which it was agreed that toxic, persistent substances that are liable to bioaccumulate arising from land-based sources should be phased out; the Barcelona Convention (1993) in which it was agreed to phase out inputs from land based sources to the Mediterranean of organohalogen compounds with these characteristics; the International Joint Commission of the Great Lakes (IJC) has called on the US and Canada to begin a phase-out of all chlorine and chlorinated organic chemicals as industrial feedstocks (IJC 1992, IJC 1994).programmeme An emergency strategy Because there is already a large burden of dioxins in the global environment which will persist for many years, aggressive measures must be implemented if the exposure of the human population is to be significantly decreased. Since all uses of chlorine and chlorinated organic chemicals are suspected of dioxin formation at one or more points in their lifecycle, the phase-out of dioxins necessitates the phase-out of all chlorine chemistry. Implementating a programme to phase-out dioxin releases from industry should be based on the following principles: * Zero means zero. Dioxin releases from industry must be eliminated, not simply reduced. The current environmental burden will take years to decrease because of the persistence of these chemicals and their continual recycling throughout the environment. Given the current health threat it would be wholly inappropriate for environmental regulatory bodies and governments to permit the release of any additional dioxin into the environment. * Pollution prevention, not control. The use of pollution control devices, filters, treatment systems and disposal methods such as burning or burying simply shifts chemicals from one environmental medium to another or delays their release until a later date. Therefore to achieve zero releases of dioxin from industry, attention should be focused preventing the release of dioxin by changing the industrial processes and feedstocks that result in its formation. * Address all dioxin sources. US EPA, WHO and other environmental regulatory bodies, must address all known industrial sources of dioxin in order to bring future releases of dioxin to zero. Research should be initiated to identify unknown and suspected sources. * Set priorities for dioxin elimination. Since dioxin is associated with the many uses of chlorine in industry, eliminating dioxins will require substantial technical and economic conversion. Timetables should be set which prioritise the largest dioxin producing sectors and for those sources for which alternatives are already available. A moratorium should be placed on new dioxin permits, so that no new permits are issued and existing ones are modified to include timetables for reduction and eventual elimination of dioxin releases. Major sources of dioxins which should be urgently considered include incineration, pulp and paper production, use and production of PVC and uses and manufacturer of chlorinated aromatic chemicals. Secondary actions to phase out other chlorine uses include phasing out chlorinated solvents and chlorine related pesticides, chlorine use in metallurgy and inorganic processes utilising chlorine. Although phasing out dioxin sources will require substantial investment in some sectors, most of the alternative products provide economic benefits in terms of increased employment, improved efficiency, decreased expenses for chemical procurement, waste disposal, liability and remediation, and the elimination of social costs associated with damage to health and the environment. Technological and economic transformation may be difficult to implement and it is essential that workers and communities should not bear the economic burden of these changes. The phasing out of dioxins should therefore be guided by a democratic transition programme to protect, compensate and provide future opportunities for workers and communities affected by the conversion. ********************************************************** 3. Dioxins in the environment and in humans Chemical characteristics of dioxins The term dioxin represents a class of chlorinated tricyclic, almost planar, aromatic ethers that exhibit similar chemical and physical characteristics. The number of chlorine atoms in these compounds varies between 1 and 8 resulting in a possible 75 polychlorinated dibenzo-p-dioxins (PCDDs) and 135 polychlorinated dibenzofurans (PCDFs). The most toxic congeners (members of this group of chemicals) have chlorine atoms in positions 2,3,7 and 8. Due to the stability and persistence of these congeners, they are ubiquitous in the environment and in humans. 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD or TCDD, sometimes called dioxin) is the most potent of all congeners and has higher toxicity than any other synthetic molecule (Fiedler and Hutzinger 1990). The dioxins can be ranked against TCDD by a system known as Toxic Equivalents (TEQ). These factors have been internationally adopted for the PCDD/Fs as well as certain types of polychlorinated biphenyls (PCBs) (WHO 1994). Current practice adopts the international toxicity equivalence factor system although older data may be reported using 2,3,7,8-TCDD concentrations alone. Other toxicity equivalents factors are in use (see eg Maisel & Hunt 1990) and total PCDD/F concentrations may also be reported. Levels of dioxin in the environment Dioxins are very chemically stable. Some degradation occurs in the atmosphere, but in other media, particularly in soils and sediments half lives are measured in decades or centuries. Global distribution of dioxins occurs largely by atmospheric transport and to a lesser extent in oceanic currents. Due to poor solubility in water and comparatively low volatility (Friesen et al. 1990), these compounds become attracted to fine particles and are predominantly transported in suspension in air or water associated with this particulate matter, rather than being in gaseous or dissolved states (Nakano et al. 1990). For this reason dioxins are often found in high concentrations in sediments, sludges and dusts. The spreading of dioxins from industrial sources by atmospheric transport was demonstrated in a study by Kjeller et al. (1991), in which agricultural soils, stored at intervals since the 1840s were analyzed. These archived samples were collected at an agricultural research station in England from land which had never been treated with pesticides or any potential source of dioxins. The levels therefore represent background figures for an industrialised nation where input was exclusively via atmospheric deposition. The data is presented in Table 1, and clearly shows the increasing levels of dioxins over time with increasing industrialization. Table 1: Increasing background contamination of agricultural soils since the start of the chlorine industry Date Concentration (pg/g total PCDD/F) 1846 61(29) 1846 54(29) 1856 31 1893 31 1914 42 1944 62 1944 57 1956 74 1966 89 1980 81 1986 95 1986 88 1986 92 Long-range atmospheric transport of PCDD/Fs from industrial sources is widespread, even in remote areas, but localised contamination around industrialised areas is also evident. For example, in the UK average soil levels in urban areas of total PCDD/Fs (1436 ng/kg) were reported to be four times higher than in rural locations (335ng/kg) (Creaser et al. 1989). Spraying of land with pesticides notably the chlorophenoxyacetic acid products 2,4-D, 2,4,5-T and agent orange has been reported to lead to significant local contamination increases (Hutzinger et al. 1982). The manufacture and disposal of chlorophenol derivatives and associated process wastes are responsible for some of the highest recorded PCDD/F pollution levels. Similar problems are seen to be associated with trichlorophenols which as well as being precursors to the chlorophenoxy acids, have a variety of other industrial applications. The site of the Spolana Chemical Company in Czechoslovakia is one of the most highly contaminated in the world, despite the fact that production of chlorophenols only occurred for three years between 1965 and 1968. Concentrations of between 0.6 and 2000 ng/g (ppb) of 2,3,7,8-TCDD was detected in production wastes. Contamination of ground waters to 0.005ug/L has also occurred (Zemek and Kocan 1991). Studies in the UK have estimated that virtually all environmental burden (99.9%) resides in soils and sediments, with soils alone representing 95% (Harrad and Jones 1992a and 1992b). It was estimated in these studies that the annual flux and contemporary UK environmental burden of PCDD/Fs correlate reasonably well. However, only 12% of the annual flux can be fully accounted for by estimated emissions from primary sources, (eg. combustion sources). This discrepancy between annual flux and emissions has also been recognised in studies in Sweden and USA (Rappe 1990, Travis et al. 1989). Although the existence of as yet unidentified sources or gross underestimates of known sources is a possible explanation of the shortfall, it is suggested that much of this disparity may be due to secondary releases from the use and disposal of chlorophenols and the recycling of existing loading through the environment. Because of the persistence of PCDD/Fs and their continual recycling through the environment it is probable that even if primary emissions of PCDD/Fs are reduced, a decrease in environmental burdens will take some time (Harrad and Jones 1992a and 1992b). PCDD/Fs tend to accumulate in the fatty tissues of animals because they are lipophilic (ie. soluble in fats and oils). This is particularly evident in aquatic environments in benthic organisms which are continually in contact with sediments, and in filterfeeders which can take up particulate matter suspended in the water column. As contaminated animals are eaten by others, the predator absorbs a large proportion of PCDD/Fs in its prey. Thus animals at the top of food chains can accrue heavy burdens, a process known as biomagnification. At the top of aquatic food chains the impact is most apparent in fish-eating colonial birds, marine mammals and polar bears (Norstrom et al. 1990, Oehme et al. 1988, Ono et al. 1989). Effects on fish and wildlife Data from both laboratory and field studies on fish and birds have shown that dioxins and furans cause detrimental effects to both adults and young but particularly embryo mortality (Gilbertson 1989, Cook and Kuehl 1991). In fish, dioxin exposure of eggs effects embryo development causing reduced hatchability and increased mortality during the "sac fry" stage. Laboratory experiments have shown high levels of PCDD/Fs in inland waters would cause early life stage mortality in feral fish, thereby reducing fish populations (Cook and Kuehl 1991). Similarly, in birds the embryo is more sensitive to dioxins than adults. For example, since the 1960s there have been several reported declines in populations of colonial fish-eating birds in the Great Lakes in which dioxins have been associated with increased embryo mortality (Fox and Weseloh 1987). Human exposure Human exposure to dioxins can occur through ingestion, inhalation and skin absorption (Kutz and Young, 1976). Inhalation and dermal contact may be important in individuals exposed to highly contaminated materials. However, the major source of exposure to the general population is through ingestion of food. It has been estimated that food intake represents 98% of exposure to PCDD/Fs in humans with the remaining 2% coming from direct uptake of air and soil (Travis and Hattemer-Frey 1987, Theelen, 1991). Food According to studies in Canada (Birmingham et al, 1989 a and b), Germany (Furst et al, 1990), Norway (Faeden, 1991), Holland (Theelen, 1991) and the UK (MAFF, 1992a), milk and milk products represent one third of the total intake of PCDD/Fs from food. Animal fat in meat, poultry and fish also contribute a large fraction of PCDD/Fs in diet. In a recent Dutch study (Theelen et al 1993) these products represented one third of the total intake and the final third came from various types of oil and fat added to processed food items. Fish oil only represented 15% of the oils which were used for this purpose but contributed to 90-95% of PCDD/Fs in the products. Unusually high concentrations of PCDD/Fs in milk/animal fats has been reported to occur near to sources of dioxin emission from industrial processes (MAFF 1992b, Rappe & Kjeller, 1987, Liem et al. 1990). High concentrations of particle-bound PCDD/Fs tend to be deposited close to emission sources resulting in contamination of herbage. Ingestion of both soil and plants by cattle in such areas can then lead to elevated levels of PCDD/Fs in milk/animal fat. For example, a study of dioxin contamination in cow~s milk from dairy farms located in the vicinity of municipal waste incinerators in the Netherlands, revealed increased levels of PCDD/Fs of up to 13.5pg TEQ/g of milk fat from areas near Rotterdam (~Lickbaert area~) and Zaandam (Liem et al. 1990). A potential health risk to humans from consumption of milk and milk products from these areas could not be ruled out. The Dutch government proclaimed an upper limit for dioxin in milk and milk products of 6pg TEQ/g of fat and provisionally restricted the consumption of these products in the affected areas. In a recent report on dioxins in food in the Netherlands, it was concluded that reduction of local emissions from industries are necessary if any substantial reduction of human exposure is to be achieved (Theelen, 1993). Sewage sludges are relatively enriched in PCDD/Fs and applications of sludges to agricultural land used for grazing can increase human exposure. Transfer of PCDD/Fs to livestock can occur by ingestion of sludge adhered to vegetation. The degree of human exposure from this is estimated to be very dependent on the amount of adherence of sludge to the vegetation (Wild et. al. 1994). It was proposed by WHO (1990) that sewage sludge application to agricultural land should be banned where there is potential for bioaccumulation in the food chain and human exposure. Given the tendency of dioxins to persist and become redistributed through the environment, spreading of dioxin contaminated sludge should not be permitted under any circumstances. Soil and sediments The occurrence of elevated levels of TCDD in soil has been considered to be a potentially significant public health concern since 1980 (Gough, 1991). Exposure can occur through dermal contact and soil ingestion. (Paustenbach, 1992). Agricultural and building workers may have a higher exposure to dioxin from these routes than the general population. Also, children are likely to have more contact with soil both dermally and through ingestion than adults. This is an important point for consideration when setting levels for remediation of residential sites. Kimbrough et al. (1984) concluded that 1 ppb of TCDD in soil was the level at which to begin consideration of action to limit human exposure. In some instances, target levels of less than 1 ppb or less have been set in the US (EPA 1987, 1990). In industrialised areas, PCDD/Fs discharged into rivers become associated with particulate matter in the water column and reach high levels in sediments, and hence reach aquatic food chains (Evers et al 1988). Exposure to humans may therefore increased in individuals who have a pronounced consumption of fish, especially sporting fishermen (Potting 1989). Levels in human blood, tissue and breast milk In general, levels of PCDD/Fs in blood, adipose tissue and milk are higher in industrialised nations than in developing nations (WHO 1989, Schecter et al. 1991b). Within a given country average dioxin tissue levels, which reflect food intake, are usually similar in diverse geographical regions and in industrialised/rural regions of a country (Furst et al. 1992). It has been found dioxins can persist in human tissues for decades. For example, individuals exposed occupationally to high levels of 2,3,7,8-TCDD and 2,3,7,8tetrabromodibenzo-p-dioxin (TBDD) had elevated levels of these compounds in their tissues up to 34 years after exposure (Schecter and Ryan, 1988). Human milk dioxin levels are of particular concern because newborns are believed to be highly sensitive to dioxins (Vos and Luster, 1989). In industrialised countries, nursing infants can ingest far more than the WHO TDI of dioxins. For example, studies undertaken by WHO (WHO 1988, WHO/EURO 1989), on levels of dioxins in human breast milk in Europe, Japan, Canada and selected areas of USA, indicate the average daily intake for breast-fed infants between 0 and 6 months of age is estimated to be 13pg 2,3,7,8TCDD/kg body weight (or 90pg TEQ/kg body weight). However, WHO (1992) stated that the TDI should not be applied to breast-fed infants since the TDI concept for these substances is based on lifetime intake. It was estimated (WHO 1990) that intake of dioxins during a 6 month nursing period would correspond to less than 5 % of the lifetime intake based on calculations of fat accumulation in the infant. It was still recommended however, that lactating mothers should not try to lose weight intentionally since dioxins could be mobilised from fat stores. Current levels of dioxins in human milk mean that the margin of safety for breast-fed babies calculated by risk assessments in Europe is very low (Theelen, 1991). Studies carried out by WHO (WHO/EURO 1989) regarding risk assessment in infants associated with exposures through human milk concluded that "a safety margin still exists, although rather small" (WHO 1992). This becomes of greater concern because health risk assessments of dioxins do not take other chemicals into account and the effects of such compounds could be additive or synergistic for given health endpoints (Schecter, 1991). Furthermore, dioxins are also transferred to the foetus via the placenta and health risks to humans appear to be particularly great in developmental stages (Mably et al. 1992, Peterson 1993, Pluim 1993). Therefore, the reduction of environmental contamination with dioxins is essential, not only for the present population but for future generations. ************************************************************* 4. Toxicity of dioxin Public concern about effeccts of dioxin on human health was hightenend in 1976 following an accident at Seveso in Italy, when an expolosion at a chemical factory caused the release of high levels of TCDD. The most commonly reported effects on humans following this accident and other incidents of high dioxin exposure was a skin rash called chloracne. However, since then numerous animal experiments and several epidemiological studies in humans have shown that dioxin causes a wide range of health effects. How dioxin exerts toxic effects Research experiments over the last 15 years have established that most toxic effects of PCDD/Fs and some PCBs are mediated by the Ah (aromatic hydrocarbon) receptor (reviewed by Okey et al. 1994, Hanson 1991). If these chemicals bind the Ah receptor, another protein called "aryl hydrocarbon receptor nuclear transferase" also interacts with the receptor to form a complex which can bind to DNA. This complex can activate the expression of specific genes once bound to DNA (Swanson et al. 1993). For example, it can activate a gene encoding cytochrome p450 enzymes (which are enzymes involved in the activation and detoxification of chemicals in the body), and there is also evidence that it may mediate expression of several other genes including those that regulate cell differentiation and growth. Therefore, dioxins can have a broad range of effects on organisms. Different dioxin congeners bind the Ah receptor with different strengths giving rise to the variation in toxicities. Although there is evidence that the Ah receptor mechanism is involved in many different effects caused by dioxin (reviewed by Peterson 1993) there may be TCDD effects which are not Ah receptor mediated. Effects on reproduction and development and the immune system Evidence from animal experiments and from incidents of accidental or occupational exposure to dioxin in humans, has shown that dioxins have detrimental effects on development, reproduction and immune system function. More disturbingly, recent experiments indicate that dioxins can have an effect on the levels of certain hormones and enzymes, and cells of the immune system, at levels (body burdens), which are at, or near to, levels currently found in the human population in industrialised countries. For example, the level of a male reproductive hormone called testosterone is affected by chronic (long-term) exposure to dioxin at low levels in animals, and after occupational exposure in humans. Another effect is the induction of some enzymes, such as the liver enzyme (cytochrome p450), the function of which is to detoxify chemicals in the body. Although the clinical significance of these effects in humans is yet to be evaluated, it is possible they could cause changes in metabolism which could lead to adverse health effects. Changes in levels of certain cells of the immune system have been found to occur in animals after chronic low dose exposure to dioxin. Members of the population who are exposed to a higher levels of dioxin, for example, from a high fish or sea mammal consumption, are more at risk of having adverse effects from dioxin (EPA 1994). According to a draft of the EPA~s reassessment: "Subtle changes in enzyme activity indicating liver changes, in levels of circulating reproductive hormones in males, in reduced glucose tolerance potentially indicative of risk of diabetes, and in cellular changes related to immune function suggest the potential for adverse impacts on human metabolism, reproductive biology, and immune compe- tence at or within one order of magnitude of average background body burden levels" Individuals at the high end of the general population range may be experiencing some of these effects. Some more highly exposed members of the population may be at risk for frankly adverse effects including developmental toxicity, reduced reproductive capacity based on decreased sperm counts and potential for increased fetal death, higher probability of experiencing endometriosis, reduced ability to withstand an immunological challenge and others." (EPA 1994) Developmental toxicity In fish, birds and mammals, exposure to dioxins causes toxic effects to the developmental stages of life. Recent laboratory studies suggest that altered development may be among the most sensitive TCDD endpoints. Because developmental toxicity following exposure to dioxins occurs across these three vertebrate classes, it is likely to occur in man (Peterson 1993). Dioxins have been reported to cause developmental toxicity in mammals including decreased growth, structural malformations, functional alterations and prenatal mortality. Functional alterations are the most sensitive including effects to the male reproductive system and male reproductive behaviour in rats, and neurobehavioural effects in monkeys (Peterson, 1993). Similar toxic effects on development have been reported in humans exposed to high levels of dioxins. Effects of accidental/occupational exposure In humans, exposure to dioxins during gestation can result in toxicity to the developing foetus with effects including foetal death, structural malformations, organ dysfunction and growth retardation. In the Yusho and/or Yu-Cheng incidents in Japan and Taiwan, in which people consumed rice oil contaminated with PCDD/Fs and PCBs, all of these outcomes occurred (Kuratsune 1989, Rogan 1989, Yamashita and Hayashi 1985). Increased incidence of prenatal mortality and low birth weight were observed suggesting foetal growth retardation had occurred (Peterson 1993). Organ dysfunction involving the central nervous system (CNS) characterised by developmental delay or psychomotor delay, including impairment of intellectual development, was reported to occur in children born to mothers exposed in the Yu-Cheng incident (Rogan Fet al. 1988, Hsu et al.). Although it can be concluded from these studies that transplacental exposure to PCDD/Fs and PCBs can affect CNS function postnatally, it is difficult to delineate which congeners were responsible (Peterson 1993). This is because some PCB congeners exert their effects through the Ah receptor, but other non TCDD-like PCBs are thought to act through a different mechanism. However, experiments in mice and monkeys do suggest that TCDD-like PCB congeners and dioxins that exert their effects through the Ah-receptor are probably involved in producing the observed postnatal neurobehavioral effects in humans (Peterson 1993). Effects on CNS function have also been reported to occur in fourteen children (7 girls and 7 boys) born between 1977 and 1983 to mothers who resided in the TCDD contaminated environment of Times Beach, Missouri, during and subsequent to pregnancy. (Cantor et al. 1993). Times Beach was contaminated with TCDD during the early 1970s when contaminated waste oil was sprayed on roads and at many horse arenas for dust control. Tests to evaluate brain function (neurophysiology) in these children relative to an age and sex matched population were carried out in 1992. Results showed that children exposed to TCDD in utero and post-natally exhibited abnormal brain measures (neurophysiological dysfunction), principally in the bilateral frontal lobe regions of the brain, relative to unexposed children, with females exhibiting more dysfunction than males. It is believed that abnormal frontal lobe region function can affect intellectual processes. Effects on development occur at very low doses In fish, birds, animal studies and in the Yusho/Yu-Cheng incidents, it is evident that the developing embryo/foetus is more sensitive to the dioxin-like chemicals than the adult mother. In animal experiments this foetal toxicity can occur at relatively high doses or chronic low dose exposure to dioxins. More disturbingly, animal experiments have shown that only transient exposure to relatively low levels of TCDD at critical times during gestation may be sufficient to cause irreversible disruptions in organ structure or function (Peterson 1993). An example of this is the developing male reproductive system in rats. Exposure to TCDD at a very low dose (64ng/kg) on day 15 of pregnancy has no detectable effect on the mother but decreased testosterone concentrations in male foetuses and neonates, and testes decent was delayed. The decreases in testosterone was partly responsible for effects from the exposure which extended into adulthood. These included a reduction in sex organ weights, a decrease in sperm count and effects on sexual behaviour. It was concluded that the male reproductive system in rats is highly sensitive to in utero and lactational TCDD exposure, and appears to be more sensitive to such exposure than any other organ or organ system in rats studied to date (Mably et al. 1991, 1992). Studies of effects on humans at current background levels In humans, of particular concern are studies which have recently been conducted on current background exposure to dioxins and PCBs. Inuits from Arctic Quebec have a relatively high body burden of PCDD/Fs and PCBs due to their diet, largely from sea mammals and fish. A study was undertaken to test whether in utero exposure to PCDD/Fs and PCBs effected the health status of Inuit newborns. Results showed that newborn males were born smaller than females. The height of the males correlated negatively with levels of both PCDD/Fs and PCBs found in breast milk fat, and female height correlated positively (Dewailley et al. 1993a). Since this observation is consistent with studies on animals and with children born to exposed mothers in the Yu-Cheng incident (Rogan et al. 1988), it suggests that levels of PCDD/Fs and PCBs currently found in these people do effect newborns by exposure in utero. In a separate study, Inuit neonates up to one year old were reported to have increased episodes of acute otitis media (infected inflammation of the middle ear) which correlated statistically with levels of PCDD/Fs and PCBs found in breast milk. The results implied that the increased acute otitis media episodes were possibly due to deficiencies in the immune system caused by exposure to PCDD/Fs and PCBs (Dewailly et al. 1993b). Dioxins influence thyroid hormone status in animals (Henry et al, 1987) which may influence the maturation of the CNS and have consequences for psychomotor development (Birrell et al, 1983). A recent study on healthy breast-fed babies has been conducted in the Netherlands. It revealed that dioxins transferred by the placenta and by breast milk caused changes in thyroid hormone concentrations and that this was most likely due to interference with the thyroid regulatory system (Pluim et al, 1993). Reproductive Toxicity In sexually mature laboratory mammals, effects of TCDD on the reproductive system have only been observed at relatively high doses, which are usually toxic to the animal. The most sensitive signs of reproductive toxicity in male and female mammals is a decrease inspermatogenesis (sperm count) and the ability to conceive and carry a pregnancy (Peterson,1993). Other effects include reduced testis and accessory sex organ weights, abnormal testis structure, reduced fertility, decreased testicular testosterone synthesis and other sex hormone effects. In females, reduced fertility, reduced litter sizeand effects on female gonads and menstrual cycle have been reported (Peterson 1993). WHO reported that in chemical workers who were accidently exposed to dioxins "there is no evidence that exposure to the male has resulted in abnormal reproductive effects" (WHO 1992). However, since then, accidental exposure of chemical workers to dioxins was shown to decrease testosterone levels (Egeland et al. 1994). It has been reported that sperm counts have fallen and disorders of the male reproductive tract have increased since the 1950~s (Sharpe & Skakkebaek 1993). It is possible that dioxins and other organochlorines play a part in this, with effects on individuals exposed in utero being greater than effects on exposed adults. In regard to in utero exposure of dioxin and human fertility, the study on in utero exposure in rats (Mably et al. 1991) discussed above, illustrated a worrying point. It was found that a single low dose exposure during day 15 of pregnancy caused a reduction in sperm count. Rats produce ten times as much sperm than required for fertilization, and consequently there was little effect on fertility. However, a reduction in sperm count in humans of a similar magnitude to that seen in the study on rats would be expected to decrease fertility in humans, because the number of sperm produced per ejaculation is closeto that required for fertility. It is therefore possible that more highly exposed members of the human population may be atrisk from decreased sperm count (EPA 1994). Effects on the immune system Evidence from numerous animal studies and some epidemiological studies in humans has shown that the immune system is a target for PCDD/Fs and PCBs. This is of concern because the function of the immune system is to maintain health, and if immune system functioning is impaired, this can result in an increase in the incidence of infectious diseases and some types of cancer. Evidence from animal studies suggests there are multiple cellular targets within the immune system that are altered by TCDD. Also, it appears that TCDD can indirectly effect the immune system, for example, by altering the activity of certain hormones (EPA 1994). Effects on immune system functioning have been found to occur in children born to mothers who resided in a TCDD contaminated area in Times Beach, Missouri, during and subsequent to pregnancy (Smoger et al. 1993). Tests were performed on these children at the ages of 9 to 14 years and revealed significant changes in the numbers of several types of cells involved in immune system function. The results were consistent with previous analyses of exposed human populations, and experiments on marmosets, and showed that immune deficiencies caused by in utero and post-natal exposure to TCDD may persist for 10 years or more. Liver damage and effects on immune system functioning were also documented to occur in patients following the Yu-Cheng episode in 1979 in Taiwan (Chang et al, 1981, 1982 a, b). However, a recent report has demonstrated possible effects on the immune system in humans at current background levels. This study on subjects in Sweden who had a high fish consumption from the Baltic Sea, showed that organochlorines including dioxins from the diet appeared to have an effect on a subset of lymphocytes (white blood cells) called natural killer cells (Svensson et al, 1993). However, further studies are needed for conclusive evidence. Carcinogenicity TCDD has been reported to be the most potent rodent carcinogen yet tested (Skene et al,1989). Animal studies have provided conclusive evidence thatTCDD is a multi stage carcinogen, increasing the incidence of tumours at sites distant from the site of treatment (USEPA 1994). Although animal studies provide substantial presumptive evidence of the human carcinogenicity of TCDD, actual validation must include evidence from human studies (USEPA 1991). Unfortunately, some epidemiological studies in humans have been flawed by methodological limitations such as no direct measurements of dioxins in blood or tissue to adequately characterise the dose, small sample size, lack ofa control group with a lower dioxin body burden, and an inadequate latency period after exposure for the development ofcancer (Schecter 1991). Epidemiological studies which have used relatively large sample sizes and some direct measurements of dioxin in blood or tissue to estimate exposure, monitored cancer incidence rates in cohorts of workers exposed to TCDD in the workplace (Manz et al. 1991, Zober et al.1990, Fingerhut et al. 1991 a and b). All of these studies reported an overall increase inmortality for all cancers combined and for lung cancer. Data from several studies most notably Hardell et al.(1979) have also suggested that soft tissue sarcomas may be associated with exposure to PCDD/Fs. In the studies by Fingerhut et al (1991 a & b) respiratory cancers were significantly increased and did not appear to be attributable to smoking. There was a significant overall increase in mortality (46%), (standardised mortality ratio (SMR)115; 95% confidence interval, 103 to 130) in all cancers combined compared to the control group. If respiratory system cancers were not included in this data, the overall increase was still significantly elevated (48%),(SMR 117; 95% confidence interval, 100 to 136). The study concluded that excess mortality from all cancers combined, from respiratory cancer and soft tissue sarcoma, was consistent with TCDD being a carcinogen. However, the study could not completely exclude the possible contribution from exposure to other occupational chemicals or smoking. Overall, data from epidemiological studies appear to be consistent with animal studies on the carcinogenicity of dioxin, and consequently US EPA concluded the following: "With regard to carcinogenicity, a weight-of-the evidence evaluation suggests that dioxin and related compounds are likely to present a cancer hazard to humans." While the epidemiological data alone are not yet deemed sufficient to characterize the cancer hazard of this class of compounds as being "known," the unequivocal evidence in animal studies, inferences drawn from mechanistic data, and the suggestive evidence of recent epidemiology studies support the characterization of dioxin and related compounds as likely cancer hazards.~ (US EPA 1994) With regard to risk assessment of cancer, US EPA estimates that current background exposures could cause up to 3% of all cancers in the US: "Modelling estimates suggest that current background exposures may result in upperbound population cancer risk estimates in the range of one in ten thousand (10~4) to 1 in a thousand (10~3) attributable to exposure to dioxin and related compounds". (US EPA 1994) It is becoming evident that some of the biochemical responses produced by TCDD and PCDFs are similar in humans and experimental animals, and that receptor-mediated events are critical to carcinogenic action. Lucier et al. (1991) investigated the binding capacity of several receptors and the induction of the enzyme cytochromeP-450 to these compounds in both human and rat tissue. The study showed that humans were as or more sensitive than rats to dioxin-like chemicals with respect to their biochemical markers. Such evidence, together with animal and epidemiological studies, suggests that TCDD is a human carcinogen. ************************************************************* 5. Risk Assessment For the purposes of risk assessments from PCDD/Fs there is general agreement that the international toxicity equivalence factor (I-TEF) which has been established by NATO/ CCMS should be used (Kutz et al, 1988, NATO/CCMS A6, 1989). Nevertheless, some national risk assessments and regulations still use their own Toxicity Equivalency Factors. Tolerable Daily Intake (TDI), is used in risk assessment to define maximum dose levels for lifelong daily exposure that may be regarded as being not detrimental to human health. The World Health Organisation (WHO) and Health and Welfare Canada have set a tolerable daily intake for PCDD/Fs of 10 pg dioxin or equivalents per kg body weight (pg/kg/d) (WHO/Euro 1991, Gilman et al. 1991, respectively). US EPA employs a different method of risk assessment in which an Acceptable Daily Intake (ADI) is calculated based on an increased lifetime cancer risk of one in a million. The US ADI is 6fg/kg/d (Greenlee et al. 1991), which is considerably lower (1670x) than the WHO and Canadian TDI. This clearly demonstrates the large scientific uncertainty in risk assessment methodology. The different methods adopted by US EPA, Health and Welfare Canada and WHO are discussed below. WHO and Health and Welfare Canada method of risk assessment WHO and Health and Welfare Canada use the Safety Factor Model to calculate risk assessment. This model assumes that there is a threshold or level of dioxin, (known as a no adverse effect level),(NOAEL), below which no deleterious effects on health occurs. The risk assessments made by WHO were based on general toxicological effects in various laboratory animal species, and included pro-carcinogenic liver toxicity, reproductive effects and immunotoxicity (WHO 1990, WHO 1992). Health and Welfare Canada based risk assessments on the rate of cancer formation in rats (Kociba et al. 1978), and birth defects and certain reproductive effects in rats (Murray 1979). From the risk assessment experiments both WHO and Health and Welfare Canada identified that no observable adverse effect was observed at a dose of 1000pg/kg/day. Health and Welfare Canada applied a safety factor of 100 to the no observable adverse effect level (NOAEL) (Boddington et al. 1990, Feeley and Grant 1993), which resulted in a TDI of 10pg TEQ/kg body weight/day. The purpose of the safety factor is to protect against the possibility that humans may be more sensitive to dioxin than rats, as well as the likely differences in sensitivity between people. WHO derived a TDI by using kinetic data which showed that the no effect level of 1000pg/kg/d in animals was equivalent to a dose of 100pgTEQ/kg/day in humans. Because of insufficient data based on reproductive effects in humans a safety factor of 10 was employed giving a TDI of 10pgTEQ/ kg/day [WHO 1990, WHO 1992]. This TDI has been adopted by most European governments. However, different safety factors have been applied to the NOAEL in some countries, or different risk assessment methods have been employed, resulting in several different TDIs, as shown in Table 2. Table 2: Tolerable daily intakes of dioxin in different countries COUNTRY TDI (pg TEQ/kg/day) Regulatory Agency/Reference UK 10 (MAFF 1992c) Norway 5 (35pg/kg/week) National Poisons Centre, Ministry of Social Affairs [Ahlborg 1992] Denmark 5 (35pg/kg/week) [Ahlborg 1992] Germany 1 Bundesgesundheitsamt (Federal Health Agency) (bga-pressedienst, 1992) Australia 10 (70pg/kg/week) National Health and Medical Research Council Source: Michelle Allsopp US EPA method of risk assessment US EPA uses the cancer statistical method of risk assessment, which calculates an excess lifetime risk of cancer for a given exposure to dioxin. It has been calculated that an exposure of 6 fg/kg/day 2,3,7,8-Cl4DD in humans will be associated with an increased lifetime cancer risk of one in a million. The mathematical model used by US EPA for calculating risk assessment is the Linear Multistage Model. In this model, the doseresponse curve for excess carcinogenic risk, is assumed to be linear through to dose zero. In other words, the model assumes that there is not a threshold (or level) below which no deleterious effect on health occurs. Therefore, the model does not predict a no adverse effect level (NOAEL) for dioxin. WHO and Canadian assumptions For the average North American exposure to dioxins is thought to be about 2-4 pgTEQ/kg/ day. Since this is less than the TDI, the Canadian government concludes that the average citizen is adequately protected (Gilman et al. 1991, Feeley and Grant 1993). WHO (1990) estimated exposure to dioxins for the general population was approximately 1.9 pgTEQ/kg/ day and it would therefore seem that the population was adequately protected by the TDI. However, there is not general agreement about using the safety factor model for risk assessment of dioxins, which is based on the following assumptions. * Dioxin has a threshold, a level below which no deleterious effect occurs. * Experiments used in risk assessment identified the most sensitive end-points. * Relative effects in animals and humans can be adequately compared based on average daily dose, even for persistent compounds like dioxin which accumulate in the body, rather than on body burden. Recent research, mostly carried out by US EPA in the reassessment of dioxin, indicates that these assumptions are not consistent with current data. Therefore, the TDI proposed by WHO and Health and Welfare Canada may not be valid. The three assumptions are discussed below. 1. Dioxin thresholds? WHO and Health and Welfare Canada argue that dioxin has a threshold because (a) no observed adverse effect levels have been identified in risk assessment experiments (b) TCDD is a tumour promotor and does not damage DNA (c) effects of TCDD are mediated by a receptor (WHO 1992, Feeley and Grant 1993). (a) A NOAEL of 1000pg/kg/d was identified in experiments by WHO and Health and Welfare Canada from risk assessment experiments. In such experiments, if a large dose of a chemical increases the rate of a toxic effect relative to unexposed control animals, but a small dose does not, it is possible that a threshold does exist for this outcome. However, it is also possible that the effect really does occur at lower levels but the experiment is not powerful enough to detect it. For example, if the true probability for developing cancer at a certain dose is 1 in 1000 and only 50 animals are tested, it is unlikely that an increased number of tumours will be observed. Thus a lack of an observed increase in an effect by itself, does not mean there is a threshold. Proper interpretation of NOEALs therefore requires an understanding of the underlying biology and the limitations of experimental protocols. Unfortunately, some regulatory agencies presume the existence of thresholds, even when biological knowledge is limited.. (b) Carcinogens that directly damage DNA in cells resulting in the formation of cancer are known as genotoxic agents. It was previously thought that if a genotoxic agent damaged the DNA in a single cell in such a way that the control mechanism for cell replication was changed, the cell could multiply unchecked and could eventually lead to the formation of cancer. Although in principle this appears to be true, it is not quite this simple because there is now evidence that cancer is a multi-stage process involving several mutations and multiple mechanisms governing selective growth of these genetically altered cells (Barret et al. 1992). Since very small doses of a DNA-damaging agent are potentially capable of causing mutations in cells which could lead to cancer, many regulatory bodies prudently assume that DNA-damaging carcinogens have no threshold. Cancer can also be caused by agents that do not directly damage DNA, (non-genotoxic agents), but instead increase the growth rate of normal and/or abnormal (mutated) cells, thereby increasing the risk of tumour development. Such agents are known as tumour promoters. Experimentally, a tumour promoter is defined as something which does not cause cancers by itself but instead greatly enhances the number of tumours initiated by another agent. Unlike DNA-damaging carcinogens, promoters are often presumed to possess thresholds. Dioxins are non-genotoxic and appear to act as tumour promoters. (reviewed by Shu et al. 1987). WHO and Health and Welfare Canada argue that because dioxin is a tumour promoter and is therefore not directly mutagenic, it will exhibit a threshold. However, this view is over~simplistic because it is thought there are multiple mechanisms responsible for tumour promotion (Barret et al. 1992). Although TCDD is not directly mutagenic, it may cause indirect damage to DNA because it can induce enzymes capable of converting certain other compounds into DNA-reactive forms. (Huss et al 1994, Webster 1994). Also, TCDD has been shown to transform certain human cells grown in cell culture into cancerous forms (Yang et al. 1992). Overall, the evidence indicates that it is not necessarily true that a threshold exists for dioxin just because it is classed as a tumour promoter. This argument is upheld by US EPA who therefore use the LMM model for risk assessment in which the dose-response curve for excess carcinogenic risk is assumed to be linear through to dose zero, and does not assume a that there is a threshold. (c) It is generally accepted that effects of TCDD are mediated by a receptor, (Ah~receptor). If dioxin binds to a receptor it can trigger different biochemical responses, which can be monitored experimentally. It has been suggested that dioxin must bind a minimum number of receptors for any effect to occur, i.e. that doses below this threshold will have no biochemical consequences. However, recent studies from the EPAs reassessment on the response of certain biochemical endpoints in animals exposed to dioxin, showed no indication of a threshold (Tritscher et al 1992, Portier et al 1993, Birnbaum 1993a). For example, ~No evidence for low dose nonlinearity was observed for sensitive biochemical responses~. (Birnbaum 1993a) This means that even at very low doses of dioxin, a level was not found where no biochemical response was detected. In another study, Portier (1993) reported: "[Our findings] are consistent with the knowledge that for some receptor-mediated responses, there is a proportional relationship between receptor occupancy and biological response, even at low ligand concentrations. The results presented here illustrate that a threshold for the biological effects of TCDD exposure cannot be assumed simply on the basis that dioxin response is receptor- mediated. If TCDD- mediated effects on cytochrome p-450 induction or epidermal growth factor ( EGF) receptor binding are reliable surrogates for toxicity or toxicity is induced by similar mechanisms, the risks from exposure to TCDD are as high or possibly higher than were estimated by the EPA using a linear model. If this is the case, there should be considerable concern for the high levels of TCDD already present in human tissues." It is unclear precisely how these biochemical endpoints relate to cancer or other toxicological endpoints, since the mechanism by which TCDD causes diseases is poorly understood. Nevertheless, these findings argue against the idea that a threshold exists for receptor-mediated events. 2. Most sensitive end-points in risk assessment The WHO and Canadian TDI depends on the assumption that the experimental effects they considered are the most sensitive endpoints. Recent research on laboratory animals, however, shows that dioxin causes effects at lower doses than previously thought. This is clearly illustrated by an experiment which found that chronic low dose exposure to TCDD in rhesus monkeys caused endometriosis (Rier et al. 1993). Endometriosis is a non-cancerous condition which occurs in humans and nonhuman primates, characterised by growth and proliferation of endometrial cells at sites outside the uterus, and is associated with chronic pain and infertility. Although its cause is unknown, endometriosis is thought to involve the immune system and regulation of hormones, both effects consistent with the action of dioxin- like compounds. A previous study has also described an association between exposure to PCBs and endometriosis (Campbell et al 1985). In the experiment by Rier et al (1993), TCDD was administered in the animals feed (from 1977-1982) at doses of 5-25 parts per trillion (ppt). The frequency of endometriosis in unexposed animals would expected to be about 30%. However, endometriosis occurred in 3 out of seven (43%) animals at a dose of 5ppt and in five out of seven (71%) at 25ppt. The lowest dose administered corresponded to approximately 130 pg/kg/d (Peterson 1994a). This is 7~8 times lower than the NOAEL of 1000 pg/kg/d proposed by WHO and Health and Welfare Canada. These preliminary results indicate that the WHO and Canadian dioxin guideline is not protective of human health because increased endometriosis was found at a dose an order of magnitude less than the assumed NOAEL. However, since the total number of animals in this study was small, further studies are in progress to confirm the findings. 3. Body burden Risk assessment proposed by the WHO and Health and Welfare Canada assumes that relative effects in rats and humans can be compared on average daily dose. However, dioxins are persistent and accumulate in the body. It is very probable that biological effects depend on concentrations present in a target organ over a critical period of time rather than on dose.The limited evidence regarding the relative response of human and animal cells exposed to the same concentration of TCDD suggests that humans are no less sensitive to certain biochemical effects than rats (Lucier et al. 1991). This conclusion was also reached by Birnbaum et al. (1993) from the EPA~s research team: ~Results in enzyme induction from both rats and mice would suggest that at current environmental levels, (1-10 pg/kg/day) people may be experiencing small but significant increases in these markers of response. Highly exposed populations may be at special risk. Since animal studies suggest that changes in hepatic enzyme induction occur at body burdens similar to those at which immunotoxicity in mice and permanent effects on the reproductive system occur in rats, it is reasonable to hypothesize that subtle effects on these parameters may be occurring in the human popula- tion. [Birnbaum 1993] Chronic low dose exposure to dioxin in animal experiments have shown adverse effects to health, for example, an effect on immune system cells in marmoset monkeys (body burden 6~8ng/kg, Neubert et al. 1992) occur at body burdens which are similar to, or within an order of magnitude of current human body burdens (estimated as 510ng/kg, EPA 1994). The body burden in monkeys which received a dose of 5ppt TCDD in diet, and appeared to cause endometriosis,(Rier et al. 1993) was 27ng/kg, and is therefore within an order of magnitude of current human body burdens (EPA 1994). Such evidence, together with other animal experiments and limited human data, indicate that current body burdens in the general population in industrialised countries are already at, or within one order of magnitude, of levels where subtle changes in enzyme activity indicating liver changes, in levels of reproduc tive hormones in males, and in cellular changes related to immune function may occur. These effects suggest there is potential for adverse impacts on human metabolism, and reproductive and immune system function (EPA 1994). This is of particular concern for more highly exposed members of the population (eg. from higher than average levels of dioxins and PCBs in diet) who may be at risk of effects on development, reproduction, such as reduced sperm count, possibly endometriosis in women, and effects on the immune system, as discussed previously in this report. The critical period of time of exposure to give an effect may be relatively short as in in utero exposure, or could be relatively long as in the development of some cancers. It is very evident from recent studies that current body burdens in women which are passed through the placenta and through breast milk to the foetus/infant are of particular concern. Future Risk Assessment US EPA is currently re-evaluating the risk of exposure to dioxin using more mechanistically based models in a so called ~biological basis for risk assessment~. Lucier et al (1993) have suggested that for all future risk assessments, modelling biological phenomena should include risk assessment at the organism level (eg) survival, the tissue level (eg) carcinogenicity and at the cellular and biochemical level. At the biochemical level, because TCDD appears to act like a potent and persistent hormone agonist, information on receptor mediated events should be included in risk assessment. Additionally, since PCDD/Fs cause reproductive/developmental effects and endocrine toxicity it is suggested that non-cancer endpoints should be included for biological based models of risk assessment. Recent studies have shown that assumptions made in the derivation of WHO and Health and Welfare Canada TDI may no be longer valid. Adverse effects have been observed at lower levels than the proposes NOAEL of 1000pg/kg/d and there is no evidence that for the existence of a threshold for dioxin effects. However, WHO have not planned reassessment of the TDI in the future. ************************************************************* 6. Formation and sources: dioxin and chlorine chemistry Dioxin is not manufactured on purpose but it is formed as an unintentional by-product in many processes in which chlorine and chlorine-derived chemicals are produced, used, and disposed of. The major primary sources of dioxins include combustion processes, especially municipal waste incinerators, and industrial processes such as the pulp and paper industry and the manufacture of organic and inorganic chlorinated chemicals. Table 3 lists known and suspected dioxin sources identified in scientific literature and by government agencies. Combustion sources Incinerators Incinerators of all kinds are known to emit dioxins. A number of studies in Europe, USA, Canada and Asia have demonstrated the presence of PCDD/Fs in fly ash and flue gas from municipal waste incinerators (Rappe & Buser 1989). Studies by Vogg and Stieglitz (1986) showed that particulate carbon can react with oxygen and inorganic chlorines with copper (II) as a catalyst to form organochlorine compounds including dioxins and furans, and concluded that this de novo synthesis is the primary source of dioxins/furans produced in municipal waste incineration. Studies have shown that fly ash from refuse incinerators and elements such as chlorine, zinc, potassium, copper and sodium may act as catalytic agents for increased dioxin formation (Hinton & Lane 1991, reviewed by Fielder et al. 1990). controlled inputs of these elements into incineration feedstock may therefore lead to unpredictable levels of dioxins being formed. 2.5 million metric tons of hazardous waste were incinerated in the USA alone in 1987 (Dempsey and Oppelt 1993). Trial burns in 1993 at the Waste Technologies Industries hazardous waste incinerator in Ohio revealed mean dioxin emissions of 1.2 grams per year (TEQ), or 13.6 ug/ton of waste burned (ENSR 1993). If all waste were burned this efficiently, atmospheric emissions alone in the US would total 34g/year. However, real-world conditions will be far from this efficient and inclusion of emissions to liquid and solid wastes will raise this estimate far higher. According to the German EPA (UBA 1990), older municipal waste incinerators (MSWI) from about 10 years ago would have emitted between 20 and 50ng TE/ Nm3; current MSWI emit approximately 8ng TE/Nm3 and state of the art ones about 1ng TE/Nm3. Hospital waste incinerators in Germany have been reported to emit 15ng TEQ/ Nm3 to 18 ng TEQ/Nm3 (reviewed by Fiedler 1990). In the US 50ng TEQ/Nm3 was detected in the flue gas of hospital waste incinerators (Hauchmann et al. 1989). Table 3: Known or suspected processes that form dioxin and related chemicals Production of chlorine gas Chlorine electrolysis with graphite electrodes Chlorine electrolysis with titanium electrodes Chemical industry~use of chlorine gas Chlorinated aromatic chemicals~manufacture (chlorobenzenes, chlorophenols, PCBs others) Pesticides Dyes Speciality chemicals Chlorinated solvents~manufacture (trichloroethylene, tetrachloroethylene, carbon tetrachloride) PVC plastic~manufacture of feedstocks (ethylene dichloride, vinyl chloride Production wastes Effluent Sludge from effluent treatment Air emissions PVC plastic products Other aliphatic organochlorines~manufacture (epichlorhydrin, hexachlorobutadiene) Some inorganic chlorides~manufacture (ferric and copper chlorides, sodium hypochlorite) Uses of chlorine gas~ other industries * Pulp and paper~chlorine bleaching Mill effluent Mill sludge Pulp and paper products Emissions from sludge incinerators Water and wastewater disinfection Refined metals~manufacture with chlorine (Ni, Mg) Use of organochlorines Manufacture of chlorine-free chemicals with chlorinated intermediates (nitrophenols, parathion, others) Degreasing/extraction with organochlorine solvents in alkaline or reactive environments Oil refining with organochlorine catalysts Use of pesticides with heat (wood treatment, etc) Iron/steel sintering with organochlorine cutting oils, solvents or plastics * Burning gasoline or diesel fuel with organochlorine additives Use of chlorine-based bleaches and detergents in washing machines and dishwashers Incineration, Recycling and Fires (primary dioxin precursor in parentheses) * Medical waste incinerators (PVC) Air emissions * Municipal waste incinerators (PVC) Air emissions Ash residues * Hazardous waste incinerators (solvents, chemical manufacturing wastes) Air emissions Ash residues Cement kilns burning hazardous waste (solvents, chemical manufacturing wastes) Air emissions Cement kiln dust Accidental fires in homes and offices (PVC) Fires at industrial factilites (PVC, PCBs, other chlorinated chemicals) Aluminium recycling/smelting (PVC) Steel and automobile recycling smelting (PVC) * Copper cable recycling/smelting (PVC) * Wood burning (pentachlorophenol wood preservatives, PVC) Environmental transformation Transformation of chlorophenols to dioxins in the environment * Addressed by the EPA in documents related to its dioxin reassessment. (Cleverly 1993, Schaum 1993). List includes sectors in which formation of dioxin or related compounds (PCBs, chlorinated dibenzofurans, and/or hexachlorobenzene) has been confirmed in chemical analyses, as well as sectors in which dioxin formation is ~known or suspected~ according to EPA (EPA 1985, PCTN 1985) or NATO (Hutzinger 1988). Sewage sludge incineration releases dioxin because of organohalogen contamination from a number of sources. Industrial tie-ins to the sewerage system can introduce a wide range of contaminants (see Johnston et al. 1993). Organochlorines such as chlorobenzenes can be introduced from domestic as well as industrial sources (Wang & Jones 1991). In the USA, the practice of chlorinating sewage may exacerbate the problem. Cement kilns have been used to incinerate hazardous wastes, but there are considerable problems with the use of this technology, most notably the lack of emissions control. Few cement kilns have the complex air pollution control devices required on hazardous waste incinerators. Handling of hazardous wastes by inexperienced personnel could also pose health and environmental threats. High levels of chlorine in the feedstock could also cause formation of alkali-halogen rings leading to clogging and poor emission control, and even under idealised test conditions, chlorinated emissions were detected (Lauber 1982). The primary motivation for the use of cement kilns to dispose of hazardous waste is economic and the imposition of stringent testing and emissions controls would probably preclude their use for this purpose. The WHO have recommended a limit of 0.1ng TEQ/m3 for all incinerators, with particular attention being given to avoiding contamination of the environment with fly ash (WHO 1990). The EC draft proposal on the incineration of hazardous waste (CEC 1993) proposes a guideline of 0.1ng/m3 until 1st January 1997 when it will become a binding limit value. Although some researchers maintain that it is possible to reduce emissions to below 0.1ng/m3 (see eg Rappe 1993), the EC draft only proposes a guideline until 1997 because standardised methods of measurement are not available. The draft directive also omits to control emissions via liquid effluents or solid wastes generated by the incinerator. In summary, it is apparent that both technical and legislative measures to prevent dioxin contamination of the environment by incineration are inadequate. Ultimately, waste minimisation and alternative destruction techniques must be implemented to replace incineration and eliminate this source of dioxin pollution. Vehicle fuels Cars are a significant source of dioxins in the urban environment and near major roads (see eg Ballschmiter et al. 1986, Bingham et al. 1989). It has been estimated that in 1988 the emissions of PCDD/Fs from cars in the UK would have been in the region of 700g/year. However, since the decline in use of leaded petrol, the emissions could have declined to half that amount. The dioxin produced by burning leaded petrol is due to the addition of dichloroethane as a scavenger (Marklund et al. 1987). Where analysis of emissions from a variety of vehicles has been carried out, leaded petrol was found to cause the highest emissions, followed by unleaded petrol. Dioxins were not detected from diesel vehicles (Markund et al. 1990). This was thought to be largely due to the far lower content of chlorine in diesel fuel (0.61ppm) compared with unleaded gasoline (14ppm) and leaded gasoline (63ppm). However, another contributing factor might be the chlorinated materials in engine lubricant oils which contain high levels of chlorine (290-310ppm) and in which PCDD/Fs are detectable (Marklund et al. 1990, Ballschmiter et al. 1986). Rotard et al. (1987) also reported ppb levels of PCDD/Fs in oils which are recycled and resold as lower priced products. This is indicative of the hazards of recycling hazardous wastes. "Natural" combustion sources Bumb et al. (1980) first reported that natural combustion sources produce PCDD/Fs. Indeed, combustion is the most likely source for preindustrial dioxin generation, which has been widely reported (see eg Ligon et al. 1989, Hashimoto et al. 1990, Czuczwa & Hites 1984 and 1985). Subsequently it has been claimed that sources such as domestic and industrial coal burning contributes a large proportion of total emissions (eg Harrad and Jones 1992b). However, studies of atmospheric emissions from coal-burning power plants did not detect any PCDD/Fs (US EPA 1987, Fiedler et al. 1990). Evidence from sediment cores from the Great Lakes (Czuczwa & Hites 1985) indicates that the increase in dioxin concentrations in the environment correlate not with coal consumption but with production of chlorinated aromatic compounds. This would argue against coal combustion being as significant a source as some researchers believe. Similarly, dioxin releases from forest fires/wood burning are probably due to contamination of the wood by phenoxyherbicides (Fiedler et al. 1990) or from resuspended material from aerial deposits (Schaum 1993). A study by the German EPA reported that ash from the combustion of chlorine-free materials such as wood, paper, and chlorine-free plastics contained undetectable or extremely low concentrations of dioxin, but after combustion of chlorinated plastics, chlorinated solvents and pesticides produced much higher concentrations dioxin were detected, as shown in Table 4, (Pohle 1991). Also, Danish studies indicate that the high dioxin emissions reported to occur from wood stoves were caused by the burning of wood impregnated with the wood preservative pentachlorophenol (Vikelsoe, 1993). [TABLE UNFORMATTED] Table 4: Dioxins in ash from the burning of organochlorines and chloride-containing materials Material # of samples Total PCDD/F (ppt) Total 2378-PCDD/F (ppt) Total PCDD/F (TEQ~ppt Writing paper ND ND ND Wood, Cotton Wool ND ND ND Polyethylene 2.9 ND <.01 Chlorine-free polychlorprene alternative 2 ND <0.1 Fir wood (w/residues?) 21.4 1.51 0.65 Coffee filters (2) 6.3-7.7 0.61-0.76 0.15-0.23 PVC plastic (10) 244-2,067 7.5-122.1 3.2-42.2 PVDC plastic 3,304 20.5 14.1 Chloro-polyethelene plastic 840 27.6 10 Chloroparaffin 1,049 14 5.3 Polychloroprene plastic 323-1,096 2.2-14.0 0.7-4.7 PVC flooring production samples 352-1,847 25.7-38.4 8.2-14.5 PVC window-frame production samples 7.5-969 22.3-52.1 8.8-18.1 PVC cables (w/coppper) 669-2,970 26.2-129.4 11.4-52.6 PVC cables (w/o copper) 416-843 20.7-46.6 7.4-16.6 Other PVC products) 158-954 7.2-492 2.5-16.5 Dichloromethane 26,302 975 478 1,1,1-trichlorotheane 21,746 831 340 Tetrachloroethylene 9,72 37.9 132 Trichloroethylene 120,915 18.5 149.5 Perchloroethylene 212 1 0.4 Chlorobenzene 16,135 0.7 0.5 p-Chloronitrobenzene 190,096 29.7 21.5 o-Chloronitrobenzene 32,293 1,331.9 216 p-Chlorotoluene 1,033 ND ND Epichlorohydrin 1,532 117 36 2,4-D 178,016 168 361 Linuron (pesticide) 3,110 253 32 Industrial sources []Processes in the pulp and paper industry Large amounts of chlorine or chlorine compounds are used in the pulp and paper industry for bleaching the pulp. Fiedler et al. (1990) have reviewed the pulping process and the stages at which dioxins are released. Dioxins are present in the effluent, sludge and the pulp itself with atmospheric emissions being less significant. Increasing the amount of chlorine dioxide used compared to chlorine gas does considerably reduce dioxin emissions, but not eliminate them. WHO (1990) recommended that other bleaching processes which do not use chlorine were available and should be adopted. The aim of this would be to reduce the levels in paper products to below 1ppt (ng/kg) TEQ to minimise migration into food in contact with paper products. Alternatives to chlorine are widely accepted (Harriman & Capps 1989) and totally chorine free (TCF) technologies are now available (see eg Henricson, 1993, Kamyr 1993), which would prevent any dioxin formation. Proponents of TCF technology point out that not only will this technology virtually eliminate organohaolgen emissions, but will allow closing of the bleach plant~s water circulation, reducing the emissions of all kinds to the aquatic environment (Kukkonen 1993). []Processes in the chemical industry The following processes have been listed according to their significance of potential sources of chlorinated and/or brominated PCDD/Fs (Heindl and Hutzinger, 1986, Hutzinger et al. 1988): 1. Processes to manufacture chlorophenols and their derivatives. 2. Processes to manufacture chlorobenzenes and substituted chlor benzenes. 3. Synthesis of aliphatic chlorine compounds. 4. Methods involving chlorine-containing intermediates. 5. Inorganic chlorochemical processes. 6. Processes applying chlorinated catalysts and solvents. 7. Processes to manufacture brominated flame-retardants (brominated biphenyls, diphenylethers, etc.) Chlorinated aromatic chemicals Chlorinated aromatics are used in the production of pesticides, nylon, synthetic rubber, dyes, pharmaceuticals and speciality chemicals for plastics. The processes of aromatic chlorination and radical side chain chlorinations are associated with dioxin production, eg. chlorophenols, chloronitrobenzenes and 2,3,6-trichlorophenoxyacetic acid (Hutzinger 1988). Extremely high contamination levels are associated with chlorophenol manufacturing facilities and associated waste disposal sites worldwide, including the UK (Berryman et al. 1991), Czechoslovakia (Zemek & Kocan 1991) and the USA (Cummings et al. 1987). Synthesis of aliphatic compounds: PVC PVC production A series of European studies have clearly established that very large quantities of dioxin are formed in the production of vinyl chloride monomer, the building block for PVC. Studies found high levels of dioxins in the environment associated with the production of PVC (Evers et al. 1989 and 1993). The thermal cracking of EDC can lead to substantial waste arisings of spent copper catalyst (Norsk Hydro 1992) which is a known catalyst in the production of dioxins. In May 1994, the Swedish Environmental Protection Agency found that PVC itself contains measurable quantities of dioxins and furans (SEPA 1994). Pure PVC suspension from two Swedish PVC producers was found to contain a full range of congeners of dioxins, furans and PCBs, with total concentrations ranging from 0.86-8.69 ppt, as shown in Table 5. [TABLE NOT FORMATTED] Table 5: Dioxin in PVC plastic products Dioxin congener Sample 1~PVC plastic (ppt) Sample 2~PVC Plastic (ppt) 2,3,7,8-TCDD <0.1 0.1 1,2,3,7,8-PeCDD <0.2 0.6 1,2,3,4,7,8-HxCDD <0.4 0.3 1,2,3,6,7,8-HxCDD 0.4 1.5 1,2,3,7,8,9-HxCDD <0.4 1.1 1,2,3,4,6,7,8-HpCDD 26 5.4 2,3,7,8-TeCDF 0.6 2.7 1,2,3,4,8(7,8)-PeCDF 0.2 8.2 2,3,4,7,8-PeCDF 0.2 6.2 2,3,4,6,7,8-HxCDF <0.2 8.2 1,2,3,4,7,8(9)-HxCDF 0.1 4.9 1,2,3,6,7,8-HxCDF <0.1 4.0 1,2,3,7,8,9-HpCDF <0.4 2.4 1,2,3,4,6,7,8-HpCDF 1.0 13.0 OCDF <0.9 7.4 OCDD 55.0 6.0 3,3,4,4-TeCB 7.90 31.00 PeCV(118) 450.00 399.00 3,3,4,4,5-PeCB 1.20 18.00 PeCB (105) 58.00 68.00 3,3,4,4,5,5-HxCB <0.70 4.70 Total PCDD/F/B <604.2 594.3 Total (TEQ) 0.86 8.69 Source: Swedish EPA 1994. Two samples of pure PVC suspension from two Swedish PVC plants were analyzed. Recent analyses commissioned by Greenpeace on effluents from a PVC plant and on residues from PVC manufacture have confirmed that this industrial sector may be an important contributor of certain dioxin congeners. In the first case, effluent from a SOLVAY plant was analyzed at the University of Amsterdam. In the second case, soils samples taken from the vicinity of the oxychlorination reactor at Norsk-Hydro in Sweden were analyzed by SAL Laboratories in the UK. The results are shown in Table 6 [TABLE NOT FORMATTED] Table 6: Dioxins in effluent and soil from two ECD/VCM plants in Europe CONGENERS EFFLUENT ) WASTE (dry weight) pg/L pg/L(TEQ) ng/g ng/g (TEQ) 2,3,7,8-TCDD n/d n/d 1,2,3,7,8-PeCDD n/d -- n/d -- 1,2,3,4,7,8-HxCDD n/d -- n/d -- 1,2,3,4,6,7,8-HpCDD 21 0.2 n/d -- OCDD 125 0.1 1.70 0.001 Untargeted isomers total 5.02 2,3,7,8,-TCDF 12.0 1.2 n/d - 1,2,3,7,8,-PeCDF 7.3 0.4 n/d - 2,3,4,7,8-PeCDF 7.3 3.7 nd - 1,2,3,4,7,8-HxCDF 12.0 1.2 3.40 0.34 1,2,3,6,7,8-HxCDF 18.0 1.8 0.65 0.07 2,3,4,6,7,8-HxCDF n/d n/d - 1,2,3,7,8,9-HxCDF n/d - 1.70 0.17 1,2,3,4,6,7,8-HpCDF 228.0 2.3 17.00 0.17 1,2,3,4,7,8,9-HpCDF 24.0 0.2 n/d - OCDF 4621.0 4.6 6700.00 6.7 Untargeted isomers total 165.8 TOTAL 5068.3 15.7 TEQ 6895.20 7.5 TEQ Source: Johnston et al 1994. TEQ based on NATO-ITEQ system. TABLE 4: Results of dioxin analyses of an effluent sample from EDC/VCM manufacture and of a contaminated soil sample. The results are reported in pg/l (parts per quadrillion) and ng/g (parts per billion) dry weight respectively. Conversions to Toxic Equivalents of 2,3,7,8-TCDD are according to the EC/NATO/CCMS/i-TEQ system. In both cases, the target congeners were the seventeen chlorinated dibenzo-p-dioxins and dibenzofurans with chlorine substitution at the 2,3,7,8 positions. In the case of the waste the total nontargeted isomers comprised 5.02 ng/g PCDD and 165.8 ng/g PCDF. In both cases the samples reveal an unusually high preponderance of OCDF. The analytical report on the waste sample also observed that in addition to the target congeners, there were numerous other PCDDs and PCDFs present. Norsk-Hydro, in a press release, have claimed that the sludge landfill at Rafnes contains 18 grammes of dioxin in total. An average loading of 0.095 grammes per kilogram is estimated, while 0.42 grammes enter the landfill each year. Assuming that this is expressed in EC/NATO TEQs, and that for convenience it is comprised of 100% OCDF then this translates to totals of 18kg at a sludge concentration of 95 grammes per kilo and an annual input of 420 grammes. These are significant amounts of persistent organochlorines. A need therefore exists for a full, reported, characterisation of the content of these sludges. A full evaluation of the behaviour of the various congeners in aquatic systems is required to justify the use of TEQ systems which have been developed solely on the basis of mammalian toxicity studies. Moreover, the dioxins serve as an index of the presence of other persistent chlorinated organics in the process wastes. PVC incineration PVC also contributes to dioxin formation during incineration and has been reported as being the largest single source of chloride in hospital waste incinerators (Marrack 1988). Several reports have found a direct relationship between the amount of PVC in waste fed into an incinerator and the amount of dioxin emitted (Ozvacic 1990, Danish EPA 1993). In the Netherlands, separation of refuse effectively removes most of the organic chlorides in combustible materials such as food and wood wastes, leaving PVC as the only major source of chlorine. Reducing PVC feed has been found to result in significant decreases in dioxin emissions (Boerekamps Kanter 1993). Based on these findings the Dutch Environment Ministry concluded: ~These new experiments by the University of Leiden demonstrate clearly a relation between the content of PVC in household waste and dioxin formation in waste incinerators. On the basis of these experiments there is no reason to reconsider present policies regarding PVC applications: the main feature of this policy is that PVC applications for which no feasible system of recycling and reuse can be established the use of more environmentally sound alternative materials is to be preferred~ (Netherlands Environment Ministry 1994). Recycling of PVC-containing products PVC has been identified as being responsible for dioxin formation during the recycling of copper in copper cables (Christman 1989), steel (Tysklind 1989, Aittola 1993) and aluminium, lead, zinc and steel (Aittola 1993). PVC in fires Accidental fires at PVC factories and in offices and homes result in high emissions of dioxins (Quebec Environment Ministry 1993, UBA 1992). In Germany, dioxin concentrations of up to 10,000 ng/m2 on surfaces from accidental fires with large quantities of PVC have been found and concentrations of up to 200ng/m2 after normal fires in homes and offices (Fiedler 1993). As a consequence of the release of dioxins from accidental fires, the German EPA and Ministry of Health have proposed the following: ~The use of plastics containing chlorine and bromine should be completely excluded, as far as is possible. UBA and BGA propose a ban on the use of plastics containing chlorine and bromine in apparatus susceptible to fire, in the manufacture of chip-board, as well as the labelling of plastics containing chlorine and if necessary a ban of the use of PVC in packaging.~ (UBA 1992). Synthesis of aliphatic compounds: chlorinated solvents Solvents account for approximately 10% of all chlorine production, and are a correspondingly significant source of dioxins. Dioxins occur in the manufacture, use and disposal of chlorinated solvents. PCDD/Fs have been identified in the manufacture of trichloroethylene, tetrachloroethylene and 1,2-dichloroethane in concentrations of up to 50ppt (sum PCDD/Fs). Hexachlorobenzene has been identified in carbon tetrachloride, trichloroethylene and tetrachloroethane (Rossberg 1986). Heindl et al. (1987) reported such results indicated that the synthesis of short-chain chlorinated hydrocarbons can lead to PCDD/F formation. The uses of chlorinated solvents in metal finishing synthesis/extraction with chlorinated solvents and dry cleaning produces dioxins (Drechler 1992). Dioxin concentrations of 140ppt have been identified in the distillation sludge from the use of perchloroethylene in dry cleaning (Lexen 1992). Inorganic chlorochemical processes Dioxins are formed during the production of chlorine gas by the chloralkali process. Older chlor-alkali facilities which use graphite electrodes produce large quantities of dioxin. Chlorine gas has been found to be contaminated with dioxin-like compounds in concentrations ranging from 40 to 210 ppb (Hutzinger 1988). High concentrations of dioxin of up to 650ppb total PCDD/Fs have been found in the sludges from spent graphite electrodes used in this process (Rappe 1991). It was thought that replacement of graphite electrodes with titanium electrodes would prevent dioxin formation during the chloralkali process, but recent data from Sweden indicates that in modern chlor-alkali plants, dioxins are still produced due to reaction of chlorine with trace quantities of organic materials found mainly in plastic pipes and valves (Andersson 1993). These results are presented in Table 7 Experiments by (Rappe et al. 1989) identified PCDFs but not PCDDs in drinking water which was chlorinated using chlorine gas. Additionally, the regeneration of carbon filters used to remove contaminants from drinking water results in dioxin formation since the filters accumulate organic contaminants (Hutzinger 1988). Table 7: Dioxins in waste from a modern chlor-alkali plant Congener Sludge (ng/L) Polymer waste (ng/L) 2,3,7,8-TCDD .021 .0025 Total TCDD .19 .02 2,3,7,8-TCDF .47 .69 Total TCDF 2.17 2.91 1,2,3,7,8-PeCDD .037 .016 Total PeCDD .37 .06 1,2,3,7,8-PeCDF 1.94 2.74 2,3,4,7,8-PeCDF .70 .79 Total PeCDF 7.48 7.66 1,2,3,4,7,8-HxCDD .16 .039 1,2,3,6,7,8-HxCDD .09 .008 123789-HxCDD .11 .011 Total HxCDD 1.62 .13 1,2,3,4,7,8-HxCDF 20.5 30.6 1,2,3,6,7,8-HxCDF 3.79 7.99 1,2,3,7,8,9-HxCDF 7.26 .77 2,3,4,6,7,8-HxCDF .84 .42 Total HxCDF 36.0 25.3 1,2,3,4,6,7,8-HpCDD .72 .13 Total HpCDD 1.48 .22 1,2,3,4,6,7,8-HpCDF 58.2 38.5 1,2,3,4,7,8,9-HpCDF 19.4 7.09 Total HpCDF 101 57.9 OCDD 5.45 .69 OCDF 63.4 87.1 Total TEQ 4.65 5.13 Source: University of Umeaa 1993. Samples are sludge and plastic pipe from a Swedish chlor-alkali plant using titanium electrodes TEQ values using NATO method. Processes involving chlorinated intermediates: pesticides 96% of all synthetic organic pesticides contain chlorine or are manufactured using chlorinated intermediates (CRA 1993). Consequently, dioxins would be expected to occur as by-products in the manufacture of virtually all synthetic pesticides including chlorophenols, chlorophenoxy acetic acids, dichloropropene, lindane, atrazine and simazine. Dioxins and furans have been detected in the widely used chlorophenoxy herbicide 2,4-D in the US at a concentration of 160 ppt (TEQ) (Schecter 1993). In the production of lindane from hexachlorocyclohexane (HCH) PCDD has been found in the ppm range, and Cl8DD up to 32mg/kg (reviewed by Fiedler et al. 1990). p-chloranil (tetrachlor-1,4- benzoquinone) is used as a fungicide and seed disinfectant and has been found to be highly contaminated with PCDD/Fs (Christman et al. 1989). Metallurgical processes Elemental chlorine gas used in the production of certain metals combines with any organic material present to form organochlorine by-products, including dioxins. High concentrations of dioxins and related compounds have been identified in the emissions from several types of metal processing plants. Dioxins may be formed when chlorine is used in the production of refined nickel and magnesium. Annual emissions from one magnesium plant in Norway are estimated to be several hundred grams of TEQ to water and 6 grams per year to air. Emissions from a nickel plant have been estimated at 1g/year (TEQ) to water alone. In addition high concentrations of PCDFs have accumulated in fish tissues near the nickel production facility (Oehme et al. 1989). Recently, iron and steel plants have been found to be major sources of dioxin. German analyses show that such plants may emit dioxin in their stack gas in concentrations of 3-10 ng/M3 (TEQ) (Lahl 1993). Dioxins are produced in this process because of the introduction of chlorinated chemicals such as cutting oils and solvents. Air emmissions from these plants account for 300-1000g/year (TEQ) in Germany alone (Lahl 1993). Since U.S production of steel is approximately twice as great as that in Germany, dioxin emissions may therefore be in the range of 600-2000g/year making this one of the largest known dioxin sources (U.S DOC 1993). The WHO (1990) recommended that emissions from the metal industry should be minimised by optimising technical procedures and equipment. ************************************************************* 7. Phasing out dioxins Recent laboratory and epidemiological research, much of it carried out under the auspices of the USEPA~s reassessment of the toxicity of dioxin, indicates that environmental contamination with PCDD/Fs has reached a critical level. It is becoming clear that the general human population and particularly those with a higher than average exposure to dioxin, for example from diet, are at risk from being affected by the dioxin contamination built up over a lifetime of exposure. Unborn and young children are most at risk. The size and longevity of the current global dioxin burden means that all sources of dioxin contamination must ultimately be eliminated if any significant lowering of human exposure levels is to be achieved. This will necessitate the phase-out of all chlorine manufacture and uses. A strategy to deal with toxic and persistent pollutants in the environment which is being increasingly implemented at both the national and international level is the precautionary approach (Stairs and Johnston 1991). This philosophy does not advocate that discharge of a compound should be allowed in the environment until damage to the environment has been proven, but rather it requires that materials should not be discharged unless it can be established that they will not be deleterious. It also avoids problems deriving from the limitations of our understanding of toxicology by removing the assumption that a safe level of a particular compound or compounds can be estimated. Thus industry is often not just required to restrict emissions of environmental toxins, but to reduce them to zero. The phasing out of toxic, persistent and bioaccumulative pollutants in the environment has been addressed at several international meetings: At the Third International Conference on the protection of the North Sea (1990) it was agreed that: "for substances that cause a major threat to the marine environment, and at least for dioxins, mercury, cadmium and lead, to achieve reductions between 1985 and 1995 of total inputs (via all pathways) of the order of 70% or more, provided that the use of Best Available Technology or other low waste technologies measures enables such reductions". he Paris Convention agreed in September 1992 to the following commitment under article 3 of Annex I on the Prevention and Elimination of Pollution from Landbased Sources: ~[I]t shall, inter alia, be the duty of the (Paris) Commission to draw up: (a) plans for the reduction and phasing out of substances that are toxic, persistent and liable to bioaccumulate arising from land-based sources;~ Similarly, all contracting parties to the Barcelona Convention agreed in October 1993 to phase out inputs from land-based sources to the Mediterranean of categories of known and suspected pollutants. Specifically, the following recommendations were passed (UNEP 1993): "the Contracting Parties reduce and phase out by the years 2005 inputs to the marine environment of toxic, persistent and bioaccumulative substances listed in the LBS Protocol, in particular organohalogen compounds having those characteristics" and: "to promote measures to reduce inputs into the marine environment and to facilitate the progressive elimination by the year 2005 of substances having proven carcinogenic, tetatogenic and/or mutagenic properties in or through the marine environment" Similarly, In the US, the American Public Health Association has made a Recommendation specifically on the phase out of chlorine and related compounds. In 1993, the APHA found that : ~the only feasible and prudent approach to eliminating the release and discharge of chlorinated organic chemicals and consequent exposure is to avoid the use of chlorine and its compounds in manufacturing processes~. On this basis the organisation resolved that chlorine and chlorinated organic compounds be treated as a class for phase out, with exceptions to be made only for uses that can be demonstrated to pose no significant hazard or for which no alternatives are available (APHA 1994). The International Joint Commission on the Great Lakes (IJC) has recognised the fact that dioxin formation occurs throughout the field of chlorine chemistry and that the mix of by-products formed in the life cycle of chlorine and chlorinated organic chemicals cannot be prevented or controlled. Thus the IJC has called on the US and Canada to begin a phase-out of all chlorine and chlorinated organic chemicals as industrial feedstocks (IJC 1992, IJC 1994). In the US the White House has also begun to move towards the comprehensive regulation of the field of chlorine chemistry. In its proposal for the Clean Water Act, President Clinton proposed to develop "a national strategy to reduce, substitute, or prohibit the use of chlorine and chlorinated compounds". Specifically this recommendation consisted of an 18-month and a 12-month period of policy formulation and review, with the focus on major chlorine uses, including pulp-bleaching, PVC and solvents. Immediate priorities: major dioxin sources Since all uses of chlorine and chlorinated organic chemicals are suspected of dioxin formation at one or more points in their life cycle, the phase-out of dioxins necessitates phase-out of all chlorine chemistry. In the largest dioxin producing sectors for which alternatives are available and feasible, action should be taken immediately. Those sectors that require longer implementation phases should be placed on time lines for dioxin elimination. Major sources of dioxins which need to be urgently considered include the following: 1. Incineration and other combustion sources Firstly, no permits for new combustion facilities that burn chlorinated wastes and products should be issued. Secondly, existing permits should include timetables for eliminating all dioxin releases. Finally, the addition of chlorinated compounds to fuels, including gasoline and motor oils should be immediately phased out. 2. Pulp and paper Alternative technology is available in the pulp and paper industry for bleaching (Harriman & Capps 1989). Presently, oxygen based and other non-chlorine based bleaching methods are available and in increasing use (see eg Henricson 1993, Kukkonen 1993). Chlorine use in this industry is therefore possible to avoid and should be phased out. 3. PVC A PVC phase-out programme should be established with progressive reductions towards zero for the production and use of PVC. A ban on short life PVC uses such as packaging, toys and non-essential medical supplies should be implemented immediately. All uses of PVC in areas susceptible to fire and products subject to combustion based recycling should be prioritised in time lines for phase-out. 4. Chlorinated aromatic chemicals A phase-out programme should be established, especially concentrating on open uses (ie.) pesticides and substances such as 1,4-dichlorobenzene which are in globally widespread domestic use.Products which are associated with the production of highly dioxin- contaminated wastes, such as the chlorophenols, should be prioritised. Secondary actions Phasing out all uses of chlorine and chlorinated organic chemicals involves significant economic and technological change that will require phased implementation. While immediate action should be given to priority sectors listed above, the following dioxin producing sectors should be placed on time lines for phase-out, with schedules based on magnitude of releases. Those for which alternatives are available should be phased out rapidly and research for alternatives in other applications must be given a high priority. 1. Chlorinated solvents A timetable for the phase-out of production and use of all chlorinated solvents should be established. Alternatives for chlorinated chemicals such as intermediates, catalysts and speciality chemicals should be developed. Some chlorine-free alternatives to chlorohydrin and phosgene intermediates have already been developed (Robert 1994). 2. Chlorine-related pesticides A phase-out timetable should be initiated. The National Academy of Sciences reported that farmers can adopt organic farming methods, reduce or eliminate the use of synthetic pesticides and still enhance their profits and crop yields (NAS 1989). 3. Metallurgy The use of chlorine in high temperature metallurgy should be phased out. 4. Water and wastewater treatment Alternatives to using chlorination in drinking water and sewage include ultraviolet light, ozone, hydrogen peroxide, slow sand filtration and membrane filtration. Time lines for the implementation of chlorine -free water treatment methods should be set while insuring that adequate disinfection continues. 5. Remediation A substantial quantity of dioxin and PCB contaminated materials are present in landfills, sediments and stored industrial wastes. Closed-loop methods for degradation of these materials are in many cases highly developed (eg Jain 1993, review by Picardi et al. 1991) and could be implemented far more widely. Economic implications Phasing-out dioxin sources will require substantial technological and economic transformation, as numerous products and processes are removed from production or converted to chlorine-free alternatives. Although this conversion will require substantial investment in some sectors, most of the alternative products and processes provide economic benefits in terms of increased employment, improved efficiency, decreased expenses for chemical conversionprocurement, waste disposal, liability and remediation, and the elimination of the social costs associated with damage to health and the environment. Technological and economic transformation may be difficult to implement and it is essential that workers and communities should not bear the economic burden of these changes. The phasing out of dioxins should therefore be guided by a democratic transition programme to protect, compensate and provide future opportunities for workers and communities affected by the conversion. ************************************************************* 8. References Ahlborg U.G. (1992) Risk assessment of PCDDs and PCDFs in the Nordic Countries. Toxic Substances Journal: 12 nos.2-4, 191- 196, April- December 1992, (Special issue: Tolerable Daily Intake of PCDDs and PCDFs). Aittola, J. Paasivirta, J and Vattulainen (1993) Measurements of Organochloro Compounds at a M