Introduction

1. 2,3,7,8- tetrachlorodibenzo-p-dioxin ( 2,3,7,8-TCDD or TCDD) is a member of a class of chemicals known as dioxins. The term dioxins refers to a group of chlorinated hydrocarbons comprising the polychlorinated dibenzo- p-dioxins (PCDDs) and the polychlorinated dibenzofurans (PCDFs). In September 2000, the Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT) commenced a review of the risk assessments of dioxins carried out by the World Health Organisation (WHO) (1), the Scientific Committee on Food (SCF; http://europe.eu.int/comm/food/fs/sc/scf/out78_en.pdf), and the United States Environmental Protection Agency (US-EPA; (www.epa.gov/ncea/pdfs/dioxin/. As part of this evaluation the COT asked for COC advice on the evidence concerning human cancer risks. The COT asked COC for a view on the approach suggested by the EPA as outlined in its draft risk assessment of dioxins.

2. The COC considered the available epidemiological and experimental data in 1993 when the Committee concluded "...that there was insufficient evidence for a causal link, but it would be prudent at present to regard TCDD as a possible human carcinogen." This was a similar conclusion to that reached by the WHO International Agency for Research on Cancer (IARC) in 1987 when TCDD was classified in group 2B (ie possibly carcinogenic to humans). The IARC undertook a further review of the literature in 1997 and concluded that TCDD should be considered as a definite human carcinogen (ie Group 1 carcinogen) (2). The Committee reviewed the IARC monograph in 1998 and, specifically, the critical epidemiology studies on TCDD cited in the monograph, ie those investigations which considered individuals whose exposure to TCDD occurred under industrial situations and was documented to be substantially higher than background exposures from environmental sources of TCDD (3-16). The Committee also considered the literature on animal studies and investigations of the carcinogenic mechanism of TCDD in animals as cited in the monograph and a number of papers on the toxicological mechanisms of TCDD (17-27). Members considered that TCDD was a potent carcinogen in laboratory animals. However, the information from the most heavily occupationally exposed cohorts suggested there was, at most, only a weak carcinogenic effect in these individuals. The Committee thus concluded, in 1998, that there were insufficient epidemiological and toxicological data on TCDD to conclude a causal link with cancer in humans, but that it would be prudent to consider TCDD as a "probable weak human carcinogen".

Review of epidemiology studies published since 1998

3. Epidemiology studies on the association between exposure to TCDD and other dioxins and cancer published since 1998 are predominantly updates of cohort investigations previously considered by the COC. The Committee agreed that, in general, these publications presented analysis of more data and had improved statistical power compared to previous studies (28-33). Although the studies were all of cohorts with exposures to mixtures of chemicals, the authors had attempted to model past exposure to TCDD and to investigate dose-response. An attempt had been made in some studies to make allowance for confounding factors (eg smoking) by making internal comparisons but in no case were individual smoking data available. It was noted that the approach taken in most of the studies to assess dose-response using back-extrapolation of TCDD or TEQ* levels in blood to estimate the body-burdens at the time of occupational exposure was generally the most appropriate approach that could be taken, particularly if the half-life of TCDD used in such calculations had been adjusted for body fat and age.
*TEQ = 2,3,7,8-TCDD Toxic Equivalence of a mixture of dioxins. The method of calculation is derived by multiplying the Toxic Equivalency Factor (TEF) for each dioxin congener by its mass concentration and then the product is summed to produce the TEQ of the mixture. The TEF is a measure of the potency of each dioxin congener relative to 2,3,7,8-TCDD which has been internationally agreed by regulatory authorities.

4. In reviewing the industrial cohort studies, Members agreed that the assessment of cancer incidence undertaken by Flesch-Janys et al, 1999 (28), on a possible association with breast cancer and by Lynge, 1998 (32), provided very limited data because of limitations due to small size, low power or inadequate exposure estimation. Members agreed that no definite conclusions could be reached on the basis of the ecological study of cancer incidence near to a municipal incinerator in France (33). The Committee considered that the results from the updates of the cancer mortality studies using the Hamburg (29), NIOSH (US National Institute for Occupational Safety and Health) (30), and Dutch (31) cohorts provided evidence for an excess total cancer mortality in exposed individuals in these cohorts of 13%-50%. The dose-response analyses, using estimated TCDD doses, showed significant results for total cancer mortality in all three studies. The highest increase in lung cancer mortality of 50% was documented in the Hamburg cohort (29). However, trend tests using the Hamburg cohort for lung cancer using TCDD or TEQ quartiles as estimates of exposure were not statistically significant. The Committee concluded that the back-extrapolation of exposure undertaken for the Hamburg and Dutch cohorts had been adequately undertaken (albeit on a minority of workers) but expressed reservations about the adequacy of the exposure estimate derived for the NIOSH cohort.

5. A 20-year mortality follow-up of the Seveso cohort had recently been published (34). Members noted that this cohort provided valuable information on the association between exposure to TCDD and cancer since the accident had resulted in exposure to TCDD and not a mixture of dioxins, and the exposed group included both men and women. In addition, the follow-up and documentation of this study were excellent with over 99% of the cohort traced. The authors reported a 10% increase in risk of total cancer mortality in males but not in females. Among males, there was a 30% increase in mortality from respiratory cancer. There were also significant increases in risk of mortality from lymphohaematopoietic cancers in both sexes (males 70%, females 80%). The risk of Hodgkin's disease was elevated in the first 10 years of follow-up whilst risk of non-Hodgkin's lymphoma and myeloid leukaemia were increased after 15 years. Members noted that lymphohaematopoietic cancers had not been identified in the industrial cohorts and commented that it would be important to continue to monitor the literature for evidence of these particular cancers associated with exposure to dioxins.

6. Overall the Committee agreed that the epidemiological data provided limited evidence of carcinogenicity. Since a positive association had been observed between exposure to TCDD and an increase in relative risk for total cancer mortality, a causal interpretation of these data was considered credible, but bias or confounding could not be ruled out. Members commented that the data were still too inconsistent to draw conclusions with regard to lung cancer. The Committee concurred with the view expressed in the draft EPA risk assessment that cancer risk attributable to dioxins related to lifetime exposures and there was therefore no reason to anticipate that children were at any different risk to adults.


Review of Quantitative Risk Assessment undertaken by EPA using epidemiology studies.

7. The Committee noted that the approach taken by the EPA in its draft risk assessment was consistent with the general approach outlined by the agency in its proposed guidelines for carcinogen risk assessment (35). In brief, dose-response data (based on estimated or calculated body-burdens of TCDD or TEQs) derived from the available industrial cohort epidemiological studies had been used to estimate the ED₀₁(dose level giving rise to 1% response). A linear extrapolation from the ED₀₁ to zero had been used to estimate risk at background body-burdens.

8. The Committee considered that modelling of dose-response data from the industrial cohort epidemiology studies was limited by the variable quality of the exposure estimations (ie extrapolation from a sub-cohort, or use of work history to estimate exposures in members of cohort for whom no biological monitoring data were available) and the uncertainties associated with back-extrapolating estimates of body burden. Members considered that the data from the NIOSH cohort were not adequate for dose-response modelling as blood/adipose tissue data on TCDD/TEQs were not available and thus inclusion of data from this study in the summary dose-response model limited the value of this particular analysis undertaken by the EPA. The Committee agreed that the linear extrapolation for ED₀₁ to estimate risks at background body-burdens was not acceptable in that the predicted kinetic profile of TCDD and other dioxins following occupational exposure predominantly via the skin over several decades was considerably different to that of background exposure via the diet. In addition, the available mechanistic data suggest a complex multi-step process involving receptor binding which is more likely to be consistent with a threshold-related response.

9. In conclusion, the Committee agreed that the review of cancer epidemiology studies and risk characterisation of cancer undertaken by EPA as part of its review of TCDD and related compounds was a detailed and valuable scientific assessment but the derivation of ED₀₁ and the slope factor and risk at background exposure levels were not appropriate for risk assessment.

Review of mechanism data

10. The Committee agreed that there is good evidence to assume that most of the toxic effects of dioxins were consequent to an initial binding to the Ah receptor (AHR) (36). Most of the evidence on TCDD induced gene transcription related to the CYP1A1 gene. It was now clear that the sequence of events from binding to AHR to transcription was very complex involving other transcription factors, chaperones such as HSP90 and regulatory proteins such as ARA9. Heterodimerisation of AHR with (Ah Receptor Nuclear Transfer Factor) (ARNT) within the nucleus is essential for TCDD activated AHR to induce DNA binding and transactivation. Heterodimerisation can also occur with hypoxia inducible factor 1-a (HIF1-a) and AHR repressor (36,37). There is also evidence that levels of these proteins may be regulated by cell type and activation and by stages of growth and differentiation. In addition there was some evidence to suggest that the phosphorylation status of AHR is important with regard to the mechanism of TCDD toxicity (38). Overall, the data were consistent with a complex multi-step process involving receptor binding and thus a threshold interpretation of TCDD induced carcinogenicity. Members noted that, although events leading up to gene transcription are quite well understood, there is very little information on how AHR induced gene-transcription leads to cancer. There was good in-vivo evidence to support the view that AHR was involved in the acute and chronic toxic effects of TCDD, including the fact that AHR null allele ("knockout") mice strains are very resistant to TCDD toxicity. However there was evidence from such strains of mice that, at very high doses, TCDD could produce toxic effects via other mechanisms. The biological significance of this is questionable since the doses required to produce these effects are higher than those tolerated in wild type animals. Overall, there are gaps in our understanding of how TCDD causes cancer. Unfortunately, the AHR knockout mice do not survive long enough to be used in a conventional cancer bioassay and it is therefore not possible to provide a clear answer on the role of AHR in TCDD-induced carcinogenesis.


General Discussion

11. The Committee reconsidered its 1998 statement and agreed that TCDD was a multi-site carcinogen in several species of laboratory animals. Members confirmed that, in addition to the limitations of the dose-response modelling of epidemiological data, which are discussed in the preceding sections of this statement, it was also inappropriate to undertake quantitative risk assessment for cancer by modelling the dose-response for tumour data in animals fed diets containing TCDD in view of the assumptions needed for extrapolation from high doses used in such studies to background environmental exposures and the uncertainties involved in inter-species extrapolation. Members agreed that the mechanism of carcinogenicity in animals was complex and it was not possible to make any detailed comment on the role of the AHR. Members thought that molecular studies of tumours from animals exposed to TCDD might be helpful with regard to identification of tumour promotion effects of TCDD. It was noted that prenatal treatment of rats with TCDD followed by postnatal treatment with the genotoxic carcinogen dimethylbenzanthracene resulted in an increased number of mammary gland adenocarcinomas compared to animals that had not been treated with TCDD. A proliferative effect of TCDD on the terminal end buds of mammary gland ductules was noted. The data were consistent with the hypothesis that TCDD has a tumour promoting effect (39).

12. The Committee confirmed that it was not possible to comment in detail on the role of AHR mediated gene transcription in humans with regard to cancer. Members were aware of the evidence for polymorphism of the AHR gene in humans (40-42), but agreed that the functional significance of these polymorphisms for risk of carcinogenicity had not been adequately investigated for any conclusions to be drawn at present.

13. Members considered that it was important to review all the available dose-response data from the epidemiology studies to determine whether there was an adequate margin of safety between reported dioxin body burdens associated with an increased risk of cancer in epidemiological studies and average background body burdens. Members noted that the nature and time-course for TEQ or TCDD body-burdens following average lifetime exposure, occupational exposure to dioxins for 20-30 years, or following exposure to TCDD after the Seveso accident were different and agreed with the evaluation of this aspect presented in the draft EPA risk characterisation document. In the case of background exposure via the diet, body burdens would gradually increase up to steady state levels at about 40 years of age. Occupational exposure would be associated with a substantially greater build up of dioxin body burden to a peak level then gradual elimination of dioxins following cessation of occupational exposure. The time-course following the Seveso accident would have been characterised by a rapid rise up to a peak body burden followed by gradual elimination of TCDD back to background levels. It was therefore difficult to compare these different exposure profiles.

14. The Committee noted that the average body burdens of dioxin in the general population were estimated to be 1-2 orders of magnitude lower than in the critical industrial cohort studies as suggested in the draft EPA risk assessment. However, in terms of TCDD blood lipid concentrations, background levels were estimated to be 2-3 orders of magnitude lower than in the critical industrial cohort studies at the time of last exposure, but only one order of magnitude lower than the Seveso cohort (43). Members agreed that, in view of the difficulties in selecting the appropriate metric of exposure, it was not possible to quantify the margin-of-safety risk assessment. However, Members noted that the excess cancer mortality reported in the heavily exposed industrial cohorts was small and commented that any increased risk of cancer at background levels of exposure is likely to be extremely small and not detectable by current epidemiological methods.

Conclusion

15. The COC agreed that TCDD should be regarded as a probable human carcinogen on the basis of all the available data. The Committee agreed that, although a precise mechanism for carcinogenesis in laboratory animals or humans could not be elucidated from the available information, the data (ie negative genotoxicity in standard assays, and evidence from studies of mechanisms) suggested that a threshold approach to risk assessment was likely to be appropriate. In this respect Members commented that any increased risk of cancer at background levels of exposure is likely to be extremely small and not detectable by current epidemiological methods.

July 2001

References

1. Van Leeuwen FXR and Younes MM (eds). Assessment of the health risks of dioxins: re-evaluation of the tolerable daily intake (TDI). Food Additives and Contaminants (2000), 17, 223-369.

2. IARC (1997). Monographs on the evaluation of carcinogenic risks to humans. Volume 69, Polychlorinated dibenzo-para-dioxins and polychlorinated dibenzofurans, pp 33-343. IARC, Lyon.

3. Collins JJ, Strauss ME, Levinskas GJ and Conner PR. The mortality experience of workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a trichlorophenol process accident. Epidemiology (1993), 4 (1), 7-13.

4. Fingerhut MA et al . Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. The New England Journal of Medicine (1991), 324 (4), 212-218.

5. Ott GA and Zober A. Cause specific mortality and cancer incidence among employees exposed to 2,3,7,8-TCDD after a 1953 reactor incident. Occupational and Environmental Medicine (1996), 53, 606-612.

6 Manz A, Berger J, Dwyer JH, Flesch-Janys D, Nagel S and Waltsgott, H. Cancer mortality among workers in chemical plant contaminated with dioxin. Lancet (1991), 338 (8773), 959-964.

7. Flesch-Janys D, Berger J, Gurn P, Manz A, Nagel S, Waltsgott H and Dwyer JH. Exposure to polychlorinated dioxins and furans (PCDD/F) and mortality in a cohort of workers from a herbicide producing plant in Hamburg, Federal Republic of Germany. American Journal of Epidemiology (1995), 142 (11), 1165-1175.

8. Flesch-Janys D. Erratum. American Journal of Epidemiology (1996), 144 (7), 716.

9. Swaen GMH. Letter to editor.American Journal of Epidemiology (1997), 146 (4), 362-363.

10. Becher H, Flesch-Janys D, Kauppinen T, Kogevinas M, Steindhorf K, Manz A and Wahrendorf J. Cancer mortality in German male workers exposed to phenoxy herbicides and dioxins. Cancer Causes and Control (1996), 7, 312-321.

11. Coggon D, Pannett B and Winter P. Mortality and incidence of cancer at four factories making phenoxy herbicides. British Journal of Industrial Medicine (1991), 48, 173-178.

12. Bueno de Mesquita HB, Doornbos G, van der Kuip DAM, Kogevinas M and Winklemann R. Occupational exposure to phenoxy herbicides and chlorophenols and cancer mortality in the Netherland. American Journal of Industrial Medicine (1993), 23, 289-300.

13. Kogevinas M et al. Cancer mortality in workers exposed to phenoxy herbicides, chlorophenols, and dioxins. American Journal of Epidemiology (1997), 145 (12), 1061-1075.

14. Kogevinas M et al. Soft tissue sarcoma and non-Hodgkin's lymphoma in workers exposed to phenoxy herbicides, chlorophenols, and dioxins: Two nested case-control studies. Epidemiology (1995), 6, 396-402.

15. Bertazzi PA, Zocchetti C, Guercilena S, Consonni D, Tironi A, Landi MT, and Pesatori AC. Dioxin exposure and cancer risk. A 15 year mortality study after the Seveso accident. Epidemiology (1997) 8 (6), 646-652.

16. Landi M, Needham LL, Lucier G, Moccarelli P, Bertazzi PA and Caporaso N. Concentrations of dioxin 20 years after Seveso. Lancet (1997), 349, 1811.

17. Safe S. Polychlorinated Biphenyls (PCBs), Polychlorinated Dibenzo-p-Dioxins, and related compounds: Environmental and mechanistic considerations which support the development of Toxic Equivalency Factors (TEFs). Critical Reviews in Toxicology (1990), 21 (1), 51-88.

18. Schiestl RH, Aubrecht J, Yap WY, Kandikonda S and Sidhom S. Polychlorinated biphenyls and 2,3,7,8-tetrachlorodibenzo-p-dioxin induce intrachromosomal recombination in-vitro and in-vivo. Cancer Research (1997), 57, 4378-4383.

19. Rowlands JC and Gustafsson JA. Aryl hydrocarbon receptor-mediated signal transduction. Critical Reviews in Toxicology (1997), 27 (2), 109-134.

20. Blankenship A and Matsumura F. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) causes an Ah receptor-dependent and ARNT-independent increase in membrane levels and activity of p60^Src. Environmental Toxicology and Pharmacology (1997), 3, 211-220.

21. Rininger JA, Stoffregen DA and Babish JG. Murine hepatic p53, RB and CDK inhibitory protein expression following acute 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Chemosphere (1997), 34 (5-7), 1557-1568.

22. Kharat I and Saatcioglus F. Antiestrogenic effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin are mediated by direct transcriptional interference with the liganded estrogen receptor. The Journal of Biological Chemistry (1996), 271 (18), 10533-10537.

23. Sewall CH, Lucier GW, Tritscher AM and Clark GC. TCDD-mediated changes in hepatic epidermal growth factor receptor may be critical event in hepatocarcinogenic action of TCDD. Carcinogenesis (1993), 14 (9), 1885-1993.

24. Worner W and Schrenk D. Influence of liver tumour promoters on apoptosis in rat hepatocytes induced by 2-acetylaminofluorene, ultraviolet light or transforming growth factor B1. Cancer Research (1993), 56 (6), 1271-1278.

25. Kohn MC, Sewall CH, Lucier GW and Portier CJ. A mechanistic model of effects of dioxin on thyroid hormones in the rat. Toxicology and Applied Pharmacology (1996), 165, 29-48.

26. Hays SM, Aylward LL, Karch NJ and Paustenbach DJ. The relative susceptibility of animals and humans to the carcinogenic hazard posed by exposure to 2,3,7,8-TCDD: An analysis using standard and internal measures of dose. Chemosphere (1997), 34 (5-7), 1507-1522.

27. Roberts RA, Nebert DW, Hickman JA, Richburg JH and Goldsworthy TL. Pertubation of the mitosis/apoptosis balance: A fundamental mechanism of toxicology. Fundamental and Applied Toxicology, (1997), 38, 107-115.

28. Flesch-Janys D, Becher H, Berger J, Manz A, Morgenstern I, Nagel S and Steindorf K. Epidemiological investigation of breast cancer incidence in a cohort of female workers with high exposure to PCDD/F and HCH. Organohalogen Compounds (1999), 44, 379-382.

29. Flesch-Janys D, Steindorf K, Gurn P and Becher H. Estimation of the cumulated exposure to polychlorinated dibenzo-p-dioxins/furans and standardised mortality ratio analysis of cancer mortality by dose in an occupationally exposed cohort. Environ Health Perspect (1998), 106 (suppl 2), 655-662.

30. Steenland K, Piacitelli L, Deddens J, Fingerhut M and Chang L I. Cancer, heart disease, and diabetes in workers exposed to 2,3,7,8-tetrachlorodibenzo-dioxin. J Nat Cancer Ins (1999), 91 (9), 779-786.

31. Hooiveld M, Heerderik D J J, Kogevinas M, Boffetta P, Needham L L, Patterson D G Jr and Bueno-de-Mesquita H B. Second follow-up of a Dutch cohort occupationally exposed to phenoxy herbicides, chlorophenols and contaminants. Am J Epidemiol (1998), 147, 891-901.

32. Lynge E. Cancer incidence in Danish phenoxy herbicide workers, 1947-1993. Environ Health Perspect (1998), 106 (suppl 2), 683-688.

33. Viel J-F, Arveux P, Baverel J and Cahn J-Y. Soft-tissue sarcoma and non-Hodgkin's lymphoma clusters around a municipal solid waste incinerator with high dioxin emission levels. Am J Epidemiol (2000), 152 (1), 13-19.

34. Bertazzi PA, Consonni D, Bachetti S, Rubagotti M, Baccarelli A, Zochetti C and Pesatori AC. Health effects of dioxin exposure: A 20-year mortality study. Am J Epidemiol (2001), 153 (11), 1031-1044.

35. EPA. Proposed guidelines for carcinogen risk assessment. Chemical Regulation Reporter, 4-19-1996, pp73-122.

36. Fisher JM, Jones KW, Whitlock JP Jr. Activation of transcription as a general mechanism of 2,3,7,8-tetrachlorodibenzo-p-dioxin action. Molecular Carcinogenesis (1989), 1, 216-221.

37. Mimura J, Ema M, Sogawa K, Fujii-Kuriyama Y. Identification of a novel mechanisms for regulation of Ah (dioxin) receptor function. Genes Dev (1999), 13, 20-25.

38. Berghard A, Gradin K, Pongratz I, Whitelaw M, Poellinger L. Cross-coupling of signal transduction pathways: the dioxin receptor mediates induction of cytochrome P4501A1 expression via protein kinase C mechanisms. Mol Cell Biol (1993), 13

39. Brown NM, Manzolillo PA, Zhang JX, Wang J and Lamartiniere CA. Prenatal TCDD and predisposition to mammary cancer in the rat. Carcinogenesis (1998), 19 (9), 1623-1629.

40. Carte S and Sogawa K. Ah receptor gene polymorphisms and human cancer susceptibility. Chapter 14 in Metabolic Polymorphisms and Susceptibility to Cancer, edited Vineis P et al. IARC Scientific Publication no.148, 1999. IARC, Lyon.

41. Anttila S, Xiang-Dong Lei, Elovaara E, Karjalainen A, Sun W, Vainio H and Hankinson O. An uncommon phenotype of poor inducibility of CYP1A1 in human lung is not ascribable to polymorphisms in the AHR, ARNT, or CYP1A1 genes. Pharmacogenetics (2000), 10, 741-751.

42. Smart J and Daly AK. Variation in induced CYP1A1 levels: relationship to CYP1A1, Ah receptor and GSTM1 polymorphisms. Pharmacogenetics (2000), 10, 11-24.

43. Smith AH and Lopipero P. Invited Commentary: How do the Seveso findings affect conclusions concerning TCDD as a human carcinogen? Am J Epidemiol (2001) 153, 1045-1047.

Top

copyright: © | updated: 29 July, 2002