What future for dioxins? As the United Nations Environment Programme targets dioxins under its new POP agreement and the EPA issues a major reassessment of dioxin health risks JEM takes a look at current research on one of the most controversial environmental contaminants. Dioxins have long been a focus for scientiÆc public and regulatory concern. They Ærst hit the headlines in the 1960s through their use in some forms of the herbicide Agent Orange during the Vietnam War.1 They also achieved notoriety as a result of the accidental release of 2,3,7,8-TCDD the most toxic form of dioxin at Seveso in 1976 an incident that left a large area contaminated. More recent cases have focused on food contamination. In the southern United States in 1997 chickens eggs and catÆsh were found to be contaminated with dioxins as a result of contaminated ingredients in animal feed.And in a high proÆle case in Belgium last year dioxin Box 1 Assessing dioxin exposures PCDD/PCDF exhibit biological effects commonly associated with chlorinated organic chemicals. Dioxin exposures are associated with an increased risk of severe skin lesions altered liver function and lipid metabolism. They can cause general weakness associated with dramatic weight loss changes in activity of various liver enzymes and depression of the immune system. They are also suspected of causing abnormalities of the endocrine and nervous systems. TCDD has been designated carcinogenic by the US EPA and the International Agency for Research on Cancer.For the risk assessment of complex mixtures of PCDD/PCDF concentrations of dioxin and furan congeners are presented in terms of toxicity equivalents (TEQs) of TCDD the best studied compound within the dioxin family. In this approach the toxicity of the 17 individual 2,3,7,8-substituted congeners is ranked in relation to the toxicity of 2,3,7,8-TCDD based on evidence that there is a common receptor-mediated mechanism of action for these compounds (see main text). The toxic equivalency factors (TEFs) allow analytical data for individual PCDD/PCDF congeners to be converted into a single TEQ and thus a single order of magnitude estimate of toxicity. While the TEQ approach has some limitations it is widely accepted in the international scientiÆc community and is fundamental to the evaluation of dioxins which always exist in nature as complex mixtures.In Bilthoven in 1990 experts from the World Health Organisation established a tolerable daily intake (TDI) for dioxins of 10 pg TCDD per kg21 body weight (bw). This was subsequently re-evaluated in the light of new epidemiological evidence especially on neurotoxicological development and the endocrine system. In May 1998 WHO established a new TDI of 1± 4 pg kg21 (bw) day21 which is now the recognised international standard for human exposure. Adapted from Dioxins and their effects on human health Factsheet 225 World Health Organisation This journal is # The Royal Society of Chemistry 2000 contamination also linked to animal feed was found to have affected a variety of foodstuffs including eggs meat and milk products.The resulting scandal over how the incident was investigated led to the downfall of the Belgian government.2 Dioxin is the collective term for the class of compounds made up of polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF). They are highly persistent organic pollutants (POPs) found in all environmental compartments and being fat soluble tend to accumulate in higher mammals and humans. Their resistance to degradation and semi-volatility means that they may be transported over long distances and because of their persistence dioxins released into the environment many years ago continue to contribute to current-day exposure. At their meeting in South Africa from 4th to 9th December 2000 parties to the United Nations Environment Programme (UNEP) POPs Convention Focus are expected to agree controls on dioxins and furans as part of a major new agreement to reduce and eventually eliminate toxic chemical releases.3 It includes a new Ænancial instrument the POPs Fund to help developing countries address the problems of dioxins and other POPs.This landmark agreement represents the Ærst stage towards global action against one of the most problematic environmental contaminants. New perspectives Dioxins and dioxin-like compounds that may have similar effects have been studied extensively over many years possibly more so than any other organic pollutant.The sheer volume of research Ændings is overwhelming. For instance a 1995 study of dioxin emissions from waste incineration ran to 500 pages,4 a recent analysis by EU the European Dioxin Inventory runs to 900 pages,5 and the EPA's latest review to be 89N J. Environ. Monit. 2000 2 Focus published shortly (see below) occupies a mammoth 2200 pages.6 Research has focused in particular on origins and sources environmental pathways and toxicity to human health (see Box 1). Recent scientiÆc advances have come in three main areas which together have provided important new insights on the dioxin issue. Firstly a series of national and international inventories have served to bettercharacteriseenvironmentalemissions and pathways.While showing large reductions in the overall volume of dioxin emissions these studies have also highlightedmajor changes in the signiÆcance of point versus non-point sources.Secondly the health risks of dioxins have been re-evaluated by agencies such as the World Health Organisation and the US Environmental Protection Agency (EPA) in the light of new evidence from epidemiological studies and laboratory investigations. These reviews suggest rather worryingly that dioxins are potentially more harmful than previously thought with signiÆcant health risks occurring close to currently observed background levels. Thirdly in addition to cancer and other health effects dioxins are nowsuspected of beingwithin the class of compounds known as endocrine disrupting chemicals (EDCs) responsible for affecting the hormone systems of humans and animals.In terms of riskassessment,thisaddsanewdimensionto an already complicated picture. A wealth of data Formost industrialised countries the sources and scale of dioxin releases to the environment are now fairly well characterised as a result of recent efforts to establish national inventories. These highlightÆvemainsources:combustionand incineration;metals smelting reÆning and processing; chemicalmanufacturing and processing; reservoir sources (such as old electrical equipment); and biological and photochemical sources.Of these combustion sources are by far themost signiÆcant.According toUNEP waste incineration and small-scale combustors (industrial and domestic boilers etc.) together account for around 60% of global releases to air.7 Environmentalists have singled out waste incineration for criticism claiming that the burning of plastics wastes containing chlorine contributes directly to airborne emissions.However a major evaluation undertaken in 1995 found no clear relationship between fuel chlorine 90N J. Environ. Monit. 2000 2 content and dioxin concentrations in Øue gases.4 The review concluded that the variabilityobservedindioxin emissionswas more likely to be explained by factors such as combustor design operating practices andmeasurement errors rather than by fuel chlorine input.While refuting these Ændings environmentalists have openedup asecond front,attacking waste incineration for concentrating dioxins in incinerator ashes.8 One consequence of the work on inventories has been to reveal wide discrepancies in national emissions and reduction efforts.In France for example Ægures from the Environment Ministry published earlier this year estimated an overall reduction in emissions of 65% since 1995.9 Stricter rules are set to cut incinerator emissions even further from 200 g in 1995 to an estimated 10 g by 2007. Greece on the other hand has been rather less successful in curbing emissions. While ofÆcial Ægures are unavailable Greenpeace claimed recently that the uncontrolled burning of waste either by individuals or on unregulated local landÆlls was releasing around 920 g of dioxins per year into the environment – more than the whole of France.9 In Norway estimated annual emissions have fallen by 96% over the last 15 years.9 According to the latest estimates from the Pollution Control Authority total emissions for 1999 were 22 g compared to 600 g in 1985.The Agency attributes much of this fall to the introduction of pollution control equipment at a single installation–a magnesium works near Olso. More stringent regulations the phasing out of chlorinated compounds and industrial closures also contributed. According to the EPA's latest estimates in the US dioxin releases to air water and land decreased by around 80% between 1987 and 1995 due primarily to reductions in air emissions from municipal and medical waste incinerators.6 Eventual reductions of greater than 95% are expected once existing regulations for municipal waste incinerators and medical waste incinerators take effect.Canada issued its Ærst national inventory in January 1999 which has subsequently been updated.10 Highlights included a 43% reduction in atmospheric releases compared to the base year of 1990 mainly as a result of the upgrading or closure of industrial facilities. Releases to water have been reduced by an estimated 99% through the implementation of new regulations in the paper and pulp industry. Another recent study from Canada highlighted the long-range transport of emissions.11 Analysing dioxin concentrations in Nunavut in the Canadian arctic scientists from the City University of New York estimated that between 74 and 85% of the measured pollution originated in the United States and a further 4±9% from even further away in Mexico.Only around 8±21% was contributed by Canadian sources. A changing picture These and other emission inventories show that the distribution of releases is changing fundamentally. As controls on point sources such as waste incineration and the metals industry take effect so other sources account for a greater proportion of overall emissions. In the French study for example dioxins from sources other than waste incineration and metallurgical facilities were estimated to account for around 30% of total emissions.Household wood burning industrial boilers and the illicit burning of PVC cables were highlighted as particularly signiÆcant.9 With emissions from waste incineration declining as a result of new technology and legislation data for other sources needs to be improved. For example only limited information is available for the iron and steel sector in some countries including Canada and the US. Emissions to water tend to be characterised only in terms of wastewaters from the pulp and paper industries ignoring other potential routes such as sewage sludge. Reviewing the current situation in the reporting of dioxin data UNEP noted recently that the number of national emission inventories is still limited with coverage in fewer than 20 countries.7 Furthermore current inventories focus almost exclusively on emissions to air and largely ignore emissions to water and land a point also emphasised in the EU's latest inventory of national sources.12,13 The EU report stresses the need to Ænd a balance between sources with a high potential emission (such as MSW incineration and pesticide production) and sources that are important in terms of environmental impact or exposure (such as pesticide use and sludge disposal).Reservoir sources are also potentially signiÆcant such as the ``old'' chemicals{ contained in {The main chemicals concerned are polychlorinated biphenyls used in transformers and capacitors 2,4,5-trichlorophenoxy acetic acid (2,4,5-T) and pentachlorophenol (PCP) used for wood treatment.transformers and capacitors together with landÆlls contaminated soils and sediments. Other shortcomings in data collection procedures also need to be overcome. At present for example there are no harmonised methods for generating and evaluating data for national dioxin inventories and coverage varies from country to country.7 Some countries provide estimates in terms of ranges whereas others use mean or median values to calculate annual emissions for a given source. Greater harmonisation of protocols for sampling and analysing stack emission water and soils is needed. Dioxins reassessed EPA is due to publish a comprehensive scientiÆc reassessment of dioxin exposure and human health effects by the end of the year.6 Begun in 1991 the EPA exercise represents probably the most thorough investigation ever into the environmental and health risks of a speciÆc pollutant.Drafts of the health effects and exposure documents were released in 1994 and reviewed by the EPA's Science Advisory Board (SAB) the following year. While responding favourably to most of the reassessment the SAB recommended that two chapters on dose±response modelling and risk characterisation be amended and that an additional document on toxic equivalency factors (TEFs) be developed. These three revised documents were issued for consultation in June 2000 and will be incorporated into the Ænal reassessment. EPA is also expected to publish a new dioxins strategy in response to the Ændings of the reassessment exercise.The review Ænds that the weight of evidence from epidemiological studies laboratory animals and ancillary experimental data suggests that exposure to dioxins and related Box 2 Measuring elimination The principle of Level of QuantiÆcation has been suggested as a baseline to assist in establishing a virtual elimination target. The Level of QuantiÆcation is deÆned as the lowest concentration that can be accurately measured using sensitive but routine sampling and analytical methods. Any measurement below the LoQ may not be reliably quantiÆable. LoQ is deÆned as LoQ~10s where s is estimated as s the standard deviation of replicate measurements of an analyte at a concentration near the detection limit.For a measurement at the LoQ the uncertainty is °30% (10s°3s) at the 99% conÆdence level. The 10s is to ensure that a measurement greater than this value will certainly be greater than 3s the level of detection. A region between 3s and 10s represents an area of less-certain quantiÆcation and a level above 10s is in the region of quantiÆcation. Adapted from ``Level of QuantiÆcation Determination PCDD/PCDF and Hexachlorobenzene'' Environmental Technology Centre Environment Canada 1999 compounds may result in a broad spectrum of health effects. In particular recent research has highlighted signiÆcant biological effects for TCDD in experimental systems. SpeciÆc data for other TCDD-like congeners and for human populations is largely lacking however.The report notes that a series of common biological steps are known to be necessary for most if not all of the observed effects of dioxin and related compounds in vertebrates including humans. These steps include the binding of dioxin-like compounds to a cellular protein called the ``Ah receptor'' as a precursor to biochemical cellular and tissue-level changes in normal biological processes. This binding is not sufÆcient in itself however and exposure to chemicals with a similar structure to TCDD will result in similar effects. The report concludes therefore that biological responses are related to cumulative exposures to all dioxin-like chemicals rather than to exposure to any single dioxin-like compound.ConÆrming that the major route of human exposure is through ingestion of foods the report estimates that TDI's have decreased signiÆcantly since the 1970s and that as of the mid-1990s adult TDIs for dioxins and dioxin-like PCBs averaged 70 pg TEQ day21. Average background exposure (exposure not readily identiÆable with point sources) led to body burdens in the late 1980s in the range 30±80 pg TEQ g21 lipid with a mid-point of approximately 55 pg TEQ g21 lipid. The average tissue level appears to be declining and the best estimate of current (late 1990s) average body burden for the US population is 25 pg TEQ g21 lipid. One of the report's most important Ændings is on how these general exposures relate to human health.``Given the assumption that TEQ intake values represent a valid comparison with TCDD exposure'' EPA says ``some of Focus these adverse impacts may be occurring at or within one order of magnitude of average background TEQ intake or body burden levels. As body burdens increase within and above this range the probability of occurrence as well as the spectrum of human noncancer response most likely increases''. As a result the ``margin of exposure between body burdens associated with background levels of exposure and levels where effects detectable in humans in terms of body burden TEQs is considerably smaller than previously estimated and in some cases may be 1 or even less''. Overall the report concludes that while dioxins have the potential to produce a spectrum of health effects there is currently no clear indication of increased disease in the general population attributable to dioxin-like compounds.This lack of evidence is symptomatic of the shortcomings of current scientiÆc data and tools rather than conclusive proof of no exposure effects. Factors such as the apparently low margin-of-exposure for non-cancer effects and the potential for signiÆcant cancer risks through incremental exposures suggest a need for further research into the impacts of these chemicals at or near current background levels. Towards zero Regulatory authorities in the US Canada and the EU as well as the UNEP POP Convention have set targets of virtual elimination of dioxins (and other toxic substances) from the environment.As environmental concentrations of these compounds become ever smaller this prompts the question of what actually constitutes ``elimination''. How should elimination be deÆned? How can such a target be measured? And how will we know when it has been achieved? The concept of Level of QuantiÆcation 91N J. Environ. Monit. 2000 2 Focus (LoQ) can be used as a baseline to assist in setting a virtual elimination target (see Box 2). LoQ is determined by assessing the variability (standard deviation) of repeated measurements of analytes at a concentration near the detection limit. The approach is well deÆned in environmental analysis but has only recently been applied to dioxins because of the lack of high quality low level datasets on emission sources.One analysis estimated the LoQ for PCDD/ PCDF to be 32 pg m23 TEQ (based on existing data) and 16 pg m23 TEQ (based on measurements in spiked blank trains and in Øy ashes).14 Another recent analytical advance has been the development of sampling protocols for the mono- through triisomers. 15 Typically concentrations of PCDDs/PCDFs are reported only in terms of the tetra- through octa-CDD/ CDF isomers. This is because the toxic isomers of interest are subsets of the tetra- through octa-isomers and hence commercial tests using the isotope dilution technique are only available for these isomers. However the monothrough tri-CDD/CDF isomers are important for understanding formation mechanisms and source±receptor relationships.New tests for these lower level isomers have recently been reported. Charting paths In terms of future research environmental agencies continue to place a high priority on improving our understanding of environmental pathways and exposure routes. EPA's reassessment hypothesises that the primary mechanism by which dioxinlike compounds enter ecological food chains and human diet is via atmospheric deposition.6 At present it is unclear whether this deposition represents primarily current contributions of dioxin and related compounds from all media 92N J. Environ. Monit. 2000 2 or is the result of past emissions that persist and recycle in the environment.Understanding the relationship between these two scenarios will be particularly important in assessing the relative contributions from individual point sources and the effectiveness of current or future control strategies. In the light of earlier incidents animal feedstuffs are also a major concern. Both the EPA and the EU are currently investigating the potential public health risks from animal feed. Other areas being investigated in the US include a national milk sampling network to measure the concentrations of dioxins and other POPs in milk and a national air monitoring network to periodically measure dioxins in ambient air.16 Our understanding of dioxins has come a long way in the nine years since EPA started its scientiÆc reassessment.Despite the plethora of inventories reviews and studies there is still much that we do not know however and dioxins seem set to be a feature of the research landscape for some time to come. At 2000z pages EPA's latest offering might look like the deÆnitive report on the subject but it will surely not be the last word. References 1 Dioxins and their effects on human health Factsheet 225 World Health Organisation June 1999. See www.who.int/inf-fs/en/fact225.html 2 J. Environ. Monit. 1999 1 63N. 3 For background documents on the POP Convention and latest news on the negotiations see www.chem.unep.ch/ pops/ 4 Relationship Between Chlorine inWaste Streams and Dioxin Emissions From Combustors ASME Research Committee on Industrial andMunicipalWastes American Society for Mechanical Engineers New York 1995.Available at www.ping.be/ƒ ping5859/Eng/ ChlorineASME.html 5 The European Dioxin Inventory DG Environment European Commission 1997. See http://europa.eu.int/comm/ environment/dioxin/download.htm 6 Exposure and Human Health Reassessment of 2,3,7,8- Tetrachlorodibenzo-p-dioxin (TCDD) and Related Compounds Environmental Protection Agency 2000 forthcoming. Drafts of the full report are available at www.epa.gov/ncea/pdfs/dioxin/ dioxreass.htm 7 Dioxin and Furan Inventories National and Regional Emissions of PCDD/PCDF UNEP Chemical Programme May 1999. 8 J. Environ. Monit.2000 2 12N. See also ``Greenpeace attacks PVC-dioxin study'' in ENDS Daily 22nd April 1997 www.ends.co.uk 9 J. Environ. Monit. 2000 5 79N. 10 Dioxins and Furans and Hexachlorobenzene Inventory of releases (second edition) Environment Canada September 2000. 11 Long-range Air Transport of Dioxin from North American Sources to Ecologically Vulnerable Receptors in Nunavut Arctic Canada B. Commoner et al. Center for the Biology of Natural Systems City University of New York October 2000. 12 Compilation of EU Dioxin Exposure and Health Data Summary Report DG Environment European Commission October 1999. Available at http://europa. eu.int/comm/environment/dioxin/ download.htm 13 Releases of Dioxins and Furans to Land and Water in Europe DG Environment European Commission 1999. Available at http://europa.eu.int/comm/ environment/dioxin/download.htm 14 Level of QuantiÆcation Determination PCDD/PCDF and Hexachlorobenzene Analysis & Air Quality Division Environmental Technology Centre Environment Canada 1999. 15 Extension of US EPA methods 0023A/ 8290 to include C12-labelled mono- diand tri-chlorinated dibenzo-p-dioxin and dibenzofuran standards B.K. Gullett J.V Ryan and D.Tabor EPA. 16 For details of current EPA research see www.epa.gov/ncea/projects.htm and for the latest EU research Ændings see http:// europa.eu.int/comm/environment/ dioxin/download.htm. Mike Sharpe