Dioxins and dioxin-like compounds

Mechanism of action

The aryl hydrocarbon receptor (AH receptor) is an ancient receptor, and its many functions have been revealed only recently. It is an over-600-million-year-old protein occurring in all vertebrates, and its homologs have been discovered in invertebrates and insects. It is classified as a member of the basic helix-loop-helix/Per-Arnt-Sim (bHLH/PAS) family of transcription factors, and it acts to modify transcription of a number of genes (see figure). AH-receptor activity is necessary for normal development and many physiological functions. Mice lacking the AH receptor (knockouts) are sick with cardiac hypertrophy, liver fibrosis, reproductive problems, and impaired immunology.

The AH receptor is relevant in toxicology for two very different reasons. First, it induces several enzymes important in the metabolism of foreign substances, so-called xenobiotics. These include both oxidative phase I enzymes and conjugative phase II enzymes, such as CYP1A2, CYP1B1, CYP2S1, CYP2A5, ALDH3, GSTA1, UGT1A1, UGT1A6, UGT1A7, and NQO1. This is in essence a protective function preventing toxic or carcinogenic effects of xenobiotics, but in some conditions it may also result in the production of reactive metabolites that are mutagenic and carcinogenic. This enzyme induction can be initiated by many natural or synthetic compounds, e.g., carcinogenic polycyclic hydrocarbons such as benzo(a)pyrene, several natural compounds, and dioxins. Secondly, AH receptors are involved in the activation or silencing of genes that lead to the toxic effects of high doses of dioxins. Because TCDD at high doses can influence the transcription of perhaps hundreds of genes, the genes crucial for the multitude of toxic effects of dioxins are still not known very well.

Binding of dioxin-like compounds to the AH receptor has made it possible to measure total dioxin-like activity of a sample using CALUX (Chemical Activated LUciferase gene eXpression) bioassay. The results have been comparable to TEQ levels measured by much-more-expensive gas chromatography–high-resolution mass spectrometry in environmental samples.

Toxicity

Dioxin toxicity is based on inappropriate activation of a physiologically important receptor, and therefore dose-response must be carefully considered. Inappropriate stimulation of many receptors leads to toxic outcomes, e.g. overdose of vitamin A leads to inappropriate activation of retinoid receptors resulting in (for example) malformations, and overdoses of corticosteroids or sex hormones lead to a multitude of adverse effects. Therefore, it is important to separate the effects of low doses causing activation of the receptor around the physiological range from the effects of high toxic doses. This is all the more important because of large differences in exposures even among humans. Western populations today are exposed to dioxins at doses leading to concentrations of 5 to 100 picograms/g (as TEQ in body fat), and the highest concentrations in accidental or deliberate poisonings have been 10,000 to 144,000 pg/g, leading to dramatic but not lethal outcomes.

The most relevant toxic outcomes of dioxins both in humans and animals are cancer and the developmental effects on offspring. Both have been documented at high doses, most accurately in animal experiments. As to developmental effects, there is an agreement that the present dioxin levels in many populations are not very far from those causing some effects, but there is not yet consensus on the safe level. As to cancer, there is a disagreement on how to extrapolate the risk from high toxic doses to the present low exposures.

While the affinity of dioxins and related industrial toxicants to the Ah receptor may not fully explain all their toxic effects including immunotoxicity, endocrine effects, and tumor promotion, toxic responses appear to be typically dose-dependent within certain concentration ranges. A multiphasic dose–response relationship has also been reported, leading to uncertainty and debate about the true role of dioxins in cancer rates. The endocrine-disrupting activity of dioxins is thought to occur as a downstream function of AH receptor activation, with thyroid status in particular being a sensitive marker of exposure. TCDD, along with the other PCDDs, PCDFs, and dioxin-like coplanar PCBs are not direct agonists or antagonists of hormones, and are not active in assays that directly screen for these activities such as ER-CALUX and AR-CALUX. These compounds have also not been shown to have any direct mutagenic or genotoxic activity. Their main action in causing cancer is cancer promotion. A mixture of PCBs such as Aroclor may contain PCB compounds which are known estrogen agonists but are not classified as dioxin-like in terms of toxicity. Mutagenic effects have been established for some lower-chlorinated chemicals such as 3-chlorodibenzofuran, which is neither persistent nor an AH receptor agonist.

Toxicity in animals

High doses. The symptoms reported to be associated with dioxin toxicity in animal studies are incredibly wide-ranging, both in the scope of the biological systems affected and in the range of dosage needed to bring these about. A single high-dose TCDD exposure causes cachexia that is fatal 1 to 6 weeks later. The LD₅₀ of TCDD varies wildly between species and even strains of the same species, with the most notable disparity being between the seemingly similar species of hamster and guinea pig. The oral LD₅₀ for guinea pigs is as low as 0.5 to 2 μg/kg body weight, whereas the oral LD₅₀ for hamsters can be as high as 1 to 5 mg/kg body weight. Even between different mouse or rat strains, there may be tenfold-to-thousandfold differences in acute toxicity. Many pathological findings are seen in the liver, thymus, and other organs. Some effects such as thymic atrophy are common in many species, but (for example) liver toxicity is typical in rabbits.

Low doses. Very few signs of toxicity are seen in adult animals after low doses, but developmental effects may occur at low dioxin levels, including foetal, infant, and possibly pubescent stages. Well-established developmental effects are cleft palate, hydronephrosis, disturbances in tooth development and sexual development, and endocrine effects. Surprisingly, enzyme induction, several developmental effects, and aversion to new foods occur at similar dose levels in animals that respond differently to acute high-dose toxicity. Therefore, it has been suggested that dioxin effects be divided to type I effects (enzyme induction etc.) and type II effects (lethality, liver damage, anorexia, and tumour promotion). The reason may be different requirements of the transactivation domain structure of the AH receptor for different genes. Some of these low-dose effects can in fact be interpreted as protective rather than toxic (enzyme induction, aversion to new foods).

Human toxicity

High doses

Toxicity of dioxins at high doses has been well-documented after accidents, deliberate poisonings, food-contamination episodes, and high industrial exposures. Three women in Vienna, Austria, were poisoned with large doses of TCDD in 1998. The highest concentration of TCDD in fat tissue was 144,000 pg/g, the highest ever reported in human beings. The main feature was chloracne, a serious skin disease. The victim survived, and other symptoms were modest after initial gastrointestinal symptoms and amenorrhea. Another acute incident was the deliberate poisoning of Victor Yushchenko, then presidential candidate of Ukraine, in 2004. TCDD concentration in fat was 108,000 pg/g. In this case, the most prominent symptoms were chloracne, hepatitis, and pancreatitis. The deliberate poisoning case shows that a human is not as sensitive as other animals, since Yushchenko survived a dose estimated at around 25 μg/kg.

Two serious food-contamination accidents were caused by PCB oils used in heat exchangers. The PCB oil leaked to rice bran oil consumed by thousands of people in Japan (Yusho disease 1968) and Taiwan (Yu-cheng disease 1979). The toxic effects have been attributed to dioxin-like PCBs and PCDFs. Their daily intake was up to 100,000 times higher than average intake presently. There were many skin problems, chloracne, swelling of eyelids, and hypersecretion of Meibomian glands in the eyes. Babies born to Yusho and Yu-cheng mothers were smaller than normal, and they had dark pigmentation and sometimes teeth at birth and tooth deformities. Foetal deaths and miscarriages were common.

Perhaps the best-known dioxin accident occurred in Seveso, Italy, in 1976. A tank of chlorophenols released its contents to air, including many kilograms of TCDD, and contaminated much of the city. The highest TCDD levels were found in children, up to 56,000 pg/g fat. Acute effects were limited to chloracne, although many animals such as rabbits died after eating contaminated grass. Dental aberrations were found after 25 years in persons exposed as children, and a slightly increased cancer risk was confirmed 35 years later.

In line with animal studies, developmental effects may be much more important than effects in adults. These include disturbances of tooth development and of sexual development.

An example of the variation in responses is clearly seen in a study following the Seveso disaster indicating that sperm count and motility were affected in different ways in exposed males, depending on whether they were exposed before, during, or after puberty.

In occupational settings, many symptoms have been seen, but exposures have always been to a multitude of chemicals including chlorophenols, chlorophenoxy acid herbicides, and solvents. Therefore, definitive proof of dioxins as causative factors has been difficult to obtain. By far the best-proven effect is chloracne. The suspected effects in adults are liver damage and alterations in heme metabolism, serum lipid levels, and thyroid functions, as well as diabetes and immunological effects.

Low exposures

Effects after low exposures such as from food have been difficult to prove. Levels of dioxins in contemporary populations are 5 to 20 pg/g (TEQ in fat) and 50 to 100 pg in older people or at least 1000 times lower than those in poisonings (see above). Tooth deformities have been considered plausible after long breast-feeding, when the dioxin concentrations were high in 1970s and 1980s. When the concentrations decreased during 1990s and 2000s, the effects were no longer seen. According to a study in Russia, sperm counts in 18–19-year-old men were lower when dioxin levels were higher at the age of 8 to 9 years. This was in industrial environments causing relatively high exposures to boys as well as their mothers. The contamination panel of the European Food Safety Agency (EFSA) recommended decreasing tolerable weekly intake (TWI) levels based on the Russian children study. This recommendation can be challenged, because it does not properly consider competing risks following from lost benefits of important and healthy food items such as certain fish. TWI levels are not applied for breast feeding, because benefits of breast milk are judged to be far more important than the remote risks of dioxins. A general conclusion may be that safety margins are not very great concerning developmental effects, but toxic effects are not likely at the present population levels of dioxins.

A number of cross-sectional studies have shown associations between type 2 diabetes and several POP compounds including dioxins. Such observational studies cannot prove causality; that is, there may be an association which does not prove that one is the cause of the other. The main problem is that similar associations can be found with many quite different POPs, which have only long half-lives and tendency to accumulate in lipids in common. This suggests that they may all be related to diet and obesity, which are by far the most common causes of type 2 diabetes.

There have been speculations on various effects of dioxins on endometriosis, sexual development, liver function, thyroid hormone levels, white-blood-cell levels, immune functions, and even learning and intelligence. While some of these effects might be possible after heavy exposures (like in the Seveso disaster), these claims are only based on potential exposures of population, not supported by actual measurements of dioxin concentrations. For example, absorption from bleached tampons claimed to be associated with endometriosis is insignificant compared with daily dioxin intake from food.

Carcinogenicity

Dioxins are well-established carcinogens in animal studies, although the precise mechanism is not clear. Dioxins are not mutagenic or genotoxic. The United States Environmental Protection Agency has categorised dioxin, and the mixture of substances associated with sources of dioxin toxicity, as a "likely human carcinogens". The International Agency for Research on Cancer has classified TCDD as a human carcinogen (class 1) on the basis of clear animal carcinogenicity and limited human data, and subsequently also 2,3,4,7,8-PCDF and PCB 126 as class 1 carcinogens. The mechanism is thought to be mainly promotion; that is, dioxins can accelerate the formation of tumours caused by other factors, and adversely affect the normal mechanisms for inhibiting tumour growth. Some researchers have also proposed that dioxin induces cancer progression through a very different mitochondrial pathway.

As with many toxic endpoints of dioxin, a clear dose–response relationship is difficult to establish. After accidental or high occupational exposures there is evidence on human carcinogenicity. Increases in cancer have been modest; in fact, reaching statistical significance has been difficult even after high accidental or occupational exposures like in the Yusho and Yucheng poisonings, Seveso accident, and combined occupational cohorts. Therefore, controversies on cancer risk at low population levels of dioxins are understandable. The problem with IARC evaluations is that they only assess hazard, that is, carcinogenicity at any dose. It is likely that there is a practical safe threshold for the non-genotoxic dioxins, and the present population levels do not possess any risk of cancer. There is thus some agreement that cancer risk is taken care of as well, if daily intake limits are set to protect from developmental effects. Among fishermen with high dioxin concentrations in their bodies, cancer deaths were decreased rather than increased. All this means that, in case of important beneficial food items and breast feeding, a thorough risk-benefit analysis is needed before setting limits, in order to avoid increased other risks or lost benefits.

Risk assessment

The uncertainty and variability in the dose–response relationship of dioxins in terms of their toxicity, as well as the ability of dioxins to bioaccumulate, have led WHO experts to recommend a very low tolerable daily intake (TDI) of dioxin, 1–4 pg/kg body weight per day, i.e. 7×10⁻¹¹ to 2.8×10⁻¹⁰ g per 70-kg person per day, to allow for this uncertainty and ensure public safety in all instances. Authorities have then set weekly or monthly intake levels that equal to TDIs around 2 pg/kg. Because dioxins are eliminated very slowly, the body burden accumulated during a whole lifetime is high compared with daily doses, and occasional modest exceedances of limit values do not change it much. Therefore, long-term intake is much more important than daily intake. Specifically, the TDI has been assessed to guarantee the safety of children born to mothers exposed to such a daily intake of dioxins all their lifetime prior to pregnancy. It is likely that the TDI for other population groups could be higher.

One important cause for differences in different assessments has been carcinogenicity. If the dose-response of TCDD in causing cancer is linear, then it might be a true risk. If the dose-response is of a threshold-type or J-shape, then there is little or no risk at the present concentrations. Understanding the mechanisms of toxicity better is hoped to increase the reliability of risk assessment. Recently, developmental effects have been reassessed by the Contamination Panel of the European Food Safety Agency (EFSA). They propose decreasing the tolerable weekly intake (TWI) from 14 pg/kg to 2 pg/kg. This is likely to cause another controversy before being accepted by European countries. Dioxin intake and levels in breast milk in 1970s and 1980s were 5 to 10 times higher than presently, and very few effects have been found, possibly mild developmental effects on teeth.

Toxicity equivalents

All dioxin-like compounds share a common mechanism of action via the aryl hydrocarbon receptor (AHR), but their potencies are very different. This means that similar effects are caused by all of them, but much larger doses of some of them are needed than of TCDD. Binding to the AHR as well as persistence in the environment and in the organism depends on the presence of so-called "lateral chlorines"; in case of dioxins and furans, these are chlorine substitutes in positions 2, 3, 7, and 8. Each additional non-lateral chlorine decreases the potency, but qualitatively the effects remain similar. Therefore, a simple sum of different dioxin congeners is not a meaningful measure of toxicity. To compare the toxicities of various congeners and to render it possible to make a toxicologically meaningful sum of a mixture, a toxicity equivalency (TEQ) concept was created.

Each congener has been given a toxicity equivalence factor (TEF). This indicates its relative toxicity as compared with TCDD. Most TEFs have been extracted from in vivo toxicity data on animals, but if these are missing (such as in case of some PCBs), less-reliable in vitro data have been used. After multiplying the actual amount or concentration of a congener by its TEF, the product is the virtual amount or concentration of TCDD having effects of the same magnitude as the compound in question. This multiplication is done for all compounds in a mixture, and these "equivalents of TCDD" can then simply be added, resulting in TEQ, the amount or concentration of TCDD toxicologically equivalent to the mixture.

The TEQ conversion makes it possible to use all studies on the best-studied TCDD to assess the toxicity of a mixture. This is most useful in regulatory work, but it can also be used in scientific studies. This resembles the common measure of all alcoholic drinks: beer, wine, and whiskey can be added together as absolute alcohol, and this sum gives the toxicologically meaningful measure of the total impact.

The TEQ only applies to dioxin-like effects mediated by the AHR. Some toxic effects (especially of PCBs) may be independent of the AHR, and those are not taken into account by using TEQs.

TEFs are also approximations with certain amount of scientific judgement rather than scientific facts. Therefore, they may be re-evaluated from time to time. There have been several TEF versions since the 1980s. The most recent re-assessment was by an expert group of the World Health Organization in 2005.

(T = tetra, Pe = penta, Hx = hexa, Hp = hepta, O = octa)

Controversy

Greenpeace and some other environmental groups have called for the chlorine industry to be phased out. However, chlorine industry supporters say that "banning chlorine would mean that millions of people in the third world would die from want of disinfected water". Sharon Beder and others have argued that the dioxin controversy has been very political and that large companies have tried to play down the seriousness of the problems of dioxin. The companies involved have often said that the campaign against dioxin is based on "fear and emotion" and not on science.

Human intake and levels

Most intake of dioxin-like chemicals is from food of animal origin: meat, dairy products, or fish predominate, depending on the country. The daily intake of dioxins and dioxin-like PCBs as TEQ is of the order of 100 pg/day, i.e. 1–2 pg/kg/day. In many countries, both the absolute and relative significance of dairy products and meat have decreased due to strict emission controls, and brought about the decrease of total intake. For example, in the United Kingdom, the total intake of PCDD/F in 1982 was 239 pg/day and in 2001 only 21 pg/day (WHO-TEQ). Since the half-lives are very long (for e.g. TCDD 7–8 years), the body burden will increase almost over the whole lifetime. Therefore, the concentrations may increase five to tenfold from age 20 to age 60. For the same reason, short-term higher intake (such as after food-contamination incidents) is not crucial unless it is extremely high or lasts for several months or years.

The highest body burdens were found in Western Europe in the 1970s and early 1980s, and the trends have been similar in the U.S. The most useful measure of time trends is concentration in breast milk measured over decades. In many countries, concentrations have decreased to about one tenth of those in the 1970s, and the total TEQ concentrations are now of the order of 5–30 pg/g fat The decrease is due to strict emission controls and also to the control of concentrations in food. In the US young adult female population (ages 20–39), the concentration was 9.7 pg/g lipid in 2001–2002 (geometric mean).

Certain professions such as subsistence fishermen in some areas are exposed to exceptionally high amounts of dioxins and related substances. This along with high industrial exposures may be the most valuable source of information on the health risks of dioxins.

Fate of dioxins in human body

Dioxins are absorbed well from the digestive tract if they are dissolved in fats or oils (such as in fish or meat). On the other hand, dioxins tend to adsorb tightly to soil particles, and absorption may be quite low: 13.8% of the given dose of TEQs in contaminated soil was absorbed.

The same features causing persistence of dioxins in the environment also cause very slow elimination in humans and animals. Because of low water solubility, kidneys cannot excrete them in urine as such. They must first be metabolised to more-water-soluble metabolites, but that metabolism, especially in humans, is extremely slow. This results in biological half-lives of several years for all dioxins. That of TCDD is estimated to be 7 to 8 years, and for other PCDD/Fs from 1.4 to 13 years, with PCDFs on average slightly shorter than PCDDs.

In mammals, dioxins are found mostly in fat. Concentrations in fat seem to be relatively similar, be it serum fat, adipose tissue fat, or milk fat. This permits measuring dioxin burden by analysing breast milk. Initially, however, at least in laboratory animals, after a single dose, high concentrations are found in the liver, but in a few days, adipose tissue will predominate. In rat liver, however, high doses cause induction of CYP1A2 enzyme, and this binds dioxins. Thus, depending on the dose, the ratio of fat and liver tissue concentrations may vary considerably in rodents.

Uses

Dioxins have no common uses. They are manufactured on a small scale for chemical and toxicological research, but mostly exist as by-products of industrial processes such as chlorine bleaching of paper pulp, pesticide manufacture, and combustion processes such as incineration. The defoliant Agent Orange contained trace amounts of dioxin impurities and caused severe health issues as a result. The wood preservative pentachlorophenol often contained dioxins and dibenzofurans as impurities. The Stockholm Convention banned the production and use of dioxins in 2001.