Minutes of 19th November 1998 Meeting COC/MIN/98/3
Minutes of the meeting held at 10.30am on Thursday 19 November 1998 in Room 136/7B, Skipton House, Department of Health, London SE1 6LH.
PRESENT
Chairman: Professor P G Blain |
||
Members: Professor C Cooper |
Assessors: Mr S Warren (PSD) |
Secretariat: Dr R J Fielder (Scientific) |
In Attendance: Item 4 Item 5 |
Item 6
Item 7 |
Item 8 |
AGENDA
| Item | Subject | Paragraph |
| 1 | Apologies for absence/Announcements | 1-2 |
| 2 | Minutes of the meeting held on 25 June 1998 (CC/MIN/98/2) | 3 |
| 3 | Matters arising 3.1 Conclusions from 19 March 1998 meeting |
4 5 6 7 |
| 4. | Chlorinated drinking water and cancer (other than bladder cancer) | 8-12 |
| 5. | NTP carcinogenicity bioassays with CS (CC/98/29) | 13-14 |
| 6. | Organochlorine insecticides and breast cancer | 15-20 |
| 7. | Alcohol and breast cancer (CC/98/23) | 21-27 |
| 8. | SAHSU study: municipal solid waste incinerators Histopathological and case note review |
28-32 |
| 9. | Papers for information 9.1 Trends in breast cancer incidence and mortality (CC/98/26) 9.2 COC views on transgenic tests (CC/98/27) 9.3 Additional papers on transgenic tests (CC/98/28) |
33 34 35 |
| 10. | Any Other Business | 36-38 |
| 11. | Date of next meeting: 18 March 1999 |
ITEM 1: APOLOGIES FOR ABSENCE
1. Apologies were received from Professors Chilvers, Dayan, Newbold, Dr Purchase, Dr Stemplewski (MCA), Dr Fisher (MAFF), Mr Cary (HSE) and Mrs Cameron (DETR).
Announcements
2. Members were reminded of the need to declare all interests before the discussion of each item.
ITEM 2: MINUTES OF THE MEETING HELD ON 25 June 1998 CC/MIN/98/2
3. The minutes were agreed without further amendments.
ITEM 3: MATTERS ARISING
3.1 Conclusions from the 25 June 1998 Meeting CC/98/21
4. Members were asked to ratify the conclusions from the 25 June 1998 meeting which had been circulated for comment.
Item 3 : Ozone
i) In view of the substantial comments received on this item, members agreed that the conclusions should be deferred until further evaluation could be undertaken by COC and COM at their Spring 1999 meetings.
Item 6: Chlorinated drinking water and bladder cancer CC/98/30
ii) The revised version of the draft statement made reference to an additional case-control study by Yang et al (Environmental Research 1998, A78, 1-6). Members agreed that the epidemiological evidence was more consistent regarding induction of bladder cancer and chlorination of drinking water but were not persuasive of a causal association. The Committee considered that conclusions regarding bladder cancer should be included as a part of an overall statement on chlorinated drinking water and cancer.
3.2 Coumarin
5. Members were informed that the joint COM/COC statement on coumarin was submitted to the 43rd session of the Council of Europe meeting where it was agreed that a genotoxic mechanism for the liver tumours in rats could not be ruled out. The COM/COC recommendation for additional in-vivo mutagenicity tests (bone marrow micronucleus and rat liver UDS) had also been agreed.
3.3 Chrysotile (CC/98/31)
6. Members noted the progress report from the HSE on actions resulting from the COC consideration of chrysotile-substitutes and in particular that the Health and Safety Commission had initiated a public consultation on proposals to further restrict use of chrysotile. Members were informed that similar restrictions across the EU were likely to be under consideration in the near future. The COC statement was also considered as part of the most recent review by the EC Scientific Committee on Toxicity, Ecotoxicity, and the Environment (SCTEE) whose opinion was published on 15 September 1998. The SCTEE opinion was in agreement with the conclusions reached by the COC.
3.4 Openness of Committee Proceedings
7. Based on comments received from the COT/COC/COM, a draft submission to Ministers and CMO was currently under consideration. A draft advertisement for a lay member for the COC (which would state the need for a scientific background) was also being prepared and would appear in the National and Ethnic press as well as on the Internet.
ITEM 4: CHLORINATED DRINKING WATER AND CANCER (OTHER THAN BLADDER CANCER) (CC/98/22)
8. This paper summarised 15 recently published papers, 5 case-control and 10 cohort studies, investigating associations between chlorinated drinking water and cancer at sites other than the bladder. The studies focused on cancers of the colon, rectum, stomach, kidney and pancreas. Rectal cancer (along with bladder cancer) was identified in the meta-analysis of Morris et al in 1992 (Am J Public Health, 82, 955-963) as a site showing a significantly increased relative risk (RR=1.38). The 15 studies were not directly comparable because they contained several different measures of exposure to chlorinated drinking water e.g. Trihalogenated methane (THM) levels, chlorinated versus non-chlorinated water sources, duration of time exposed to chlorinated water, and high compared to low levels of water mutagenicity.
9. Members agreed that there remained considerable inconsistency in the epidemiological data relating to risks of cancer at sites other than the bladder and that it was difficult to evaluate the impact of the studies published since 1992 because the quality of the earlier studies had not been evaluated. Comments on some of the newer studies were:
i) Hildesheim et al 1998 (Epidemiology, 9, 29-35) - this was considered to be a well-conducted case-control study that contained a large number of cases (ca 560 colon, 537 rectal, cf 1983 controls) and that had been appropriately adjusted for risk factors. There were a number of different ways of analysing the data with respect to exposure, all of which showed an increased risk for rectal, but not colon cancer. Exposure to chlorination by-products was estimated by measuring duration of lifetime residence served by chlorinated water and estimating lifetime trihalomethane (THM) exposure.
ii) Doyle et al 1997 (Am J Public Health, 87, 1168-1176) - this was also considered to be a good study that identified a clear dose response for colon cancer in post-menopausal women with increasing levels of chloroform in drinking water although no such association was found for rectal cancer. One problem with the study was the short history of water exposure for cancer cases, which was only recorded up to the date of cancer diagnosis. Members noted that diet was an important confounding factor for colon cancer.
iii) King (Health Canada, December 1995) - this study was difficult to evaluate because the data had not been presented adequately; it was noted that the results for colon and rectal cancers had not been published.
iv) Koivusalo et al 1994, 1997 (Am J Public Health 84: 1223-1228; Cancer Causes and Control 8: 192-200) - both studies used estimated mutagenicity in water, as measured by the Ames test, as a measure of exposure to chlorination by-products. The assessment of historical exposure was based on model calculations using historical information on water quality and water treatment practices. The Committee agreed that such retrospective calculations of mutagenicity in drinking water were inadequate and could not be used in the conduct of epidemiological studies as a method for estimating exposure. It was noted that the Koivusalo studies also suffered from lack of adjustment for confounders and covariants.
v) Hoff et al 1994 - (European Journal of Cancer, 1, 423-428); this study examined the exposure to total organic carbon (TOC) and chloroform but no other THM was measured. Members acknowledged that estimating exposure to chlorination by-products was difficult and that THM measurements gave only a guide to what the by-product concentration might be and did not provide information on individual exposure. This depended on the quality of the surface water abstracted, which varied considerably with season, and on the treatment processes used which may have changed over the years. Members were informed that chloroform was first detected in the early 1970s but routine monitoring data for THMs were only available in the UK since 1990; it was unlikely that there was regular monitoring for chloroform in the US before the 1980s.
10. In discussing the different findings for colon and rectal cancers, Members noted the possibility of misclassification of these tumours especially those found at the junction between the colon and rectum and suggested that it might be easier to look at the two types of tumours together. It was noted that the Hildesheim et al study had separated out the junction between the colon and rectum and that it was less likely that tumours at these locations had been misclassified in this particular investigation.
11. With regard to the animal carcinogenicity studies with disinfection by-products, the only responses supporting the epidemiological findings were tumours of the large intestine in rats obtained with certain brominated species (bromoform, bromo-dichloromethane and dibromoacetate). It was noted that intestinal tumours were unusual in rodents but that their relevance was unlikely to be significant in view of the very high doses used.
12. Members considered that a statement, which considered all of the available evidence on drinking water and cancer, should be drafted. In the conclusions, there should be a clear distinction between consistency of the evidence and magnitude of the effect. The statement should also include a discussion of the following issues, (i) the ranking of the epidemiological studies to highlight the better quality studies, (ii) a discussion of the difficulties with exposure assessment which needed to be highlighted and (iii) a discussion of biological plausibility with respect to bladder cancer and reference to diagnostic problems especially with colon and rectal cancers and the fact that the different research groups had not consistently considered this issue.
ITEM 5: NTP CARCINOGENICITY BIOASSAYS WITH 2-Chlorobenzylidene malonitrile (CS). CC/98/29
13. Members were informed that the Department of Health and the Home Office were seeking independent expert advice on the health effects that may arise from the use of CS spray as an incapacitant. Review papers were considered by the Committee on Mutagenicity of Chemicals in Food, Consumer Products and the Environment (COM) at its meeting of the 15 October and the Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT) at its meeting of the 27 October. The COM felt that for complete reassurance of the lack of in-vivo mutagenicity further data from a study to investigate genotoxicity at the initial site of contact might be required although it was recognised that negative NTP carcinogenicity bioassays would provide adequate reassurance in this regard. The COM recommended that advice of the COC be sought on the adequacy of the NTP studies.
14. Members considered the protocol to be satisfactory and the study well conducted. Members commented on the observed irritant effect on the respiratory epithelium, especially in the olfactory region, which was particularly sensitive. Members agreed that any carcinogenic effect of CS would have been expected in the olfactory region. Thus, it was concluded that the NTP carcinogenicity study provided no evidence that CS had any carcinogenic effects in adequately conducted inhalation bioassays in rats or in mice. During discussions, one member raised the possibility that repeated dermal exposure to CS might act as a promoter of skin tumours. It was agreed that this should be brought to the attention of the COT.
ITEM 6: ORGANOCHLORINE INSECTICIDES AND BREAST CANCER CC/98/24
15. A number of articles in the scientific literature and general media had suggested that prolonged exposure to certain organochlorine (OC) insecticides may cause breast cancer. The COC reviewed the available epidemiological studies in 1995 on three chemicals (DDT and isomers/metabolites, and the hexachlorocyclohexane isomers gamma-HCH (lindane) and beta-HCH. The Committee agreed in 1995, that the available evidence indicated no clear association, but had agreed that the matter should be kept under review. Since 1995, there have been a few new published epidemiological studies, but considerably more investigations of the potential oestrogenicity of these compounds. Members were told that there had also been a number of parliamentary questions and enquiries from members of the public regarding organochlorine insecticides and breast cancer. In view of the publication of new data and the continued public concern, it was felt that it was timely to update the Committee's advice.
16. Members first discussed the epidemiological studies. Essentially, there were 4 new studies, using both prospective and case-control methods. Three of these (Lopez-Carillo et al 1997, Cancer Research, 57, 3728-3732; Hunter et al, 1997, New Eng J Med, 337, 1253-1258; and van't Veer et al 1997, BMJ, 315, 81-85) investigated p, p' DDE (metabolite of DDT) and were considered to be substantive and well conducted. In all of these studies, the methods used regarding the identification of case patients and matching controls were adequate, and information on a range of breast cancer risk factors was obtained. These studies produced no evidence of an association between p, p' DDE or DDE and breast cancer. The fourth study by Guttes et al 1998 (Archives Env Cont Toxicol, 35, 140-147) was much smaller (45 cases) and comparatively less information on breast cancer risk factors was presented, although members noted that other OC insecticides such as beta-HCH and lindane, in addition to p, p' DDE, had been analysed. This study showed no convincing evidence of an association between DDT/DDE or b-HCH and breast cancer.
17. Members noted that the studies had used a variety of methods for measuring long term exposures to DDT. For example Lopez-Carillo et al measured serum levels of p,p'DDE in a very high-exposure group, Hunter et al measured DDE in plasma lipids, and van't Veer et al measured p,p'DDE aspirates of subcutaneous fat from the buttocks. Members noted that there was still some debate in the literature as to whether serum analysis was a good indicator for breast tissue but overall it was agreed that there were no concerns regarding the analytical procedures used in these studies. Members concluded that there was no evidence of an elevated relative risk of breast cancer in association with DDT; the evidence for this was now stronger than in 1995.
18. There was little epidemiological data for lindane and beta-HCH. Members noted that new analytical methods were now available that allowed the detection of lindane and other OC insecticides at very low levels down to around 28 ng/l (Stellman SD et al (1998), Cancer Epidemiology, Biomarkers and Prevention, 7, 489-496). It was agreed that such low-levels would not be significant to human health unless they were for a very potent carcinogenic compound. Members were also aware that there may also be some additional epidemiological data forthcoming in respect of DDE and beta-HCH from a large ongoing case-control study in the USA.
19. Members discussed the oestrogenic potential of these chemicals. They considered that there were good data to show that some DDT/DDE isomers and metabolites and beta-HCH had oestrogenic activity in vivo in studies using experimental animals. However, on the basis of in vitro studies, this activity was small (i.e. < 1000 x) compared to oestradiol. There was no evidence from in-vitro studies (mainly the MCF-7 assay) to suggest that lindane had direct oestrogenic activity although it would be useful to see results for lindane from other assays in order to reach a final conclusion.
20. Members agreed that a statement could be drafted based on the information presented in CC/98/24 taking into account the previous consideration in 1995, the most recent information on estrogenic activity and the available epidemiology data. The statement would also need to consider and compare the oestrogenicity of these OC insecticides with naturally occurring compounds in the diet, the potential for interaction between chemicals and the potential oestrogenicity of mixtures.
ITEM 7: ALCOHOL AND BREAST CANCER (CC/98/23)
21. Members were given the background to this item. The Government established an Interdepartmental Working Group (IDWG) to carry out a review of its sensible drinking message in the light of evidence that drinking alcohol might promote protection from Coronary Heart Disease (CHD). The COC was asked to advise on the carcinogenicity of alcoholic beverages. The COC reviewed the available epidemiological evidence for an association between alcohol and breast cancer as part of the extensive review of alcohol and cancer in 1995. The COC concluded that ..."while there is no decisive evidence that breast cancer is causally related to drinking alcohol, the potential significance, for public health, of even a weak association between alcohol and breast cancer is such that we recommend, in particular, that this matter is kept under review." There have been a number of epidemiological studies since 1995 which have further investigated the potential association between breast cancer and the consumption of alcohol and have also presented some evidence for a mechanism. It was therefore timely to reconsider the available information.
22. CC/98/23 presented the conclusions from a detailed review of the epidemiology studies on alcohol and breast cancer published since 1995 which had been prepared under contract by the Imperial College of Science, Technology and Medicine, who would be preparing some papers for the COC in the future. Three new cohort studies and 22 case-control studies investigating the association between alcohol consumption and breast cancer were evaluated. Members were told that a further paper on the evidence regarding whether or not there was a plausible mechanism by which the consumption of alcohol could lead to breast cancer would be submitted to the March 1999 meeting.
23. The Committee agreed that it was important to assess all the evidence regarding alcohol and breast cancer since the public health implications of a causal association were of major importance. Members noted that the Imperial report considered the issues of dose-response and duration of drinking, which were considered in the 1995 review, and came to similar conclusions; namely that there was evidence in these new studies of a weak association. Overall, the Imperial report proposed that new post 1995 data were consistent with the evidence considered by the COC in 1995 that there was a weak association between alcohol consumption and increased risk of breast cancer in women. The authors considered the evidence for different relative risks in pre-menopausal compared to post-menopausal women and vice-versa but were unable to find any convincing evidence regarding differences in risk. Additionally there was no definitive data regarding an effect of beverage type on relative risk and thus the authors concluded that most information pointed to an effect of alcohol itself rather than any congeners or other ingredients.
24. Members agreed that the paper produced by Imperial was a substantive detailed review that provided a descriptive narrative but agreed that a more comprehensive review of all the epidemiology data was required, particularly with respect to the quality of the papers. Members noted that the tables in the annexes could be modified to provide some of the required information such as timing of alcohol consumption.
25. Members suggested that the epidemiological papers should be quality assessed using a formal scoring system (there were published papers on how to do this). It was suggested that the system used by COMA in its report on diet and cancer might be adapted for use here. It was agreed that the secretariat should put together a proposal regarding the full assessment of the epidemiology data to address the issues raised by Members and that this would be circulated to the epidemiologists and Chair for comment.
26. Members also suggested that a formal systematic review (meta-analysis) be considered and commented that the study provided to SCOTH had been particularly helpful in assessing causality of passive smoking. It was, however, recognised that this would be a very large undertaking and the committee wished to avoid any undue delay in the COC reporting on this subject. It was agreed that the secretariat would report on the available published studies at the March 1999 meeting.
27. Members agreed that they were not in a position to finalise their conclusions on alcohol intake and breast cancer, but agreed that a holding statement could be drafted for consideration in respect of the White Paper on Health Strategy. The secretariat would circulate a draft to members with agreement by Chairman's action.
ITEM 8: SAHSU STUDY : MUNICIPAL SOLID WASTE INCINERATORS: HISTOPATHOLOGICAL AND CASE NOTE REVIEW (CC/98/25)
28. The COC first considered a report from the Small Area Health Statistics Unit (SAHSU) of a study of cancer incidence around municipal incinerators in November 1993. The results suggested a small but significant excess of liver cancer cases within 1km of incinerators. The evidence of residual confounding, particularly the finding of misdiagnosis mainly of secondary tumours among registrations and death certificates of liver cancer suggested that further investigations were required before any conclusions could be drawn. The COC recommended a histological review of the liver cancer cases to help determine whether there was an increase in primary liver cancers near incinerators.
29. The COC considered a draft report of a pilot investigation of tissue diagnosis at its 25 June 1998 meeting. Members raised a number of queries which were summarised in a letter to the authors (Annex 2), as a result, the report was re-drafted (Annex 3). The main revisions to the report were:
i) A more detailed discussion of why the relative risk estimate from the original study was retained (but not the absolute risk estimate). Because of the small numbers, the study had low power to address questions of relative risk of primary liver cancer although it could address the absolute risk. This was because, in the absence of any obvious trends in patterns of diagnosis of primary liver cancer and associated risk factors (such as cirrhosis) with distance from incinerators, it had to be assumed that any deficiencies in the registration system would affect both numerator (cases) and denominator (expected numbers) equally, leaving estimates of relative risk unchanged. By contrast, any tendency for liver cancer cases to be overestimated would give high estimates of the absolute numbers of excess cases.
ii) One extra case (case note review, no histopathological sample) had been included in the revised report.
iii) Confidence intervals for the percentage of confirmed primary liver cancer had been included.
iv) Further discussion on the angiosarcoma cases had been included. The limitations of using the national register to estimate the expected number of cases were discussed, in particular, the lack of confirmatory histopathology for 3/4 of the cases included in the database. It was noted that cases of angiosarcomas on the national register showed no evidence for geographical clustering near incinerators. The SAHSU group stated the need for caution in interpreting the apparently large relative risk estimates produced based on 2 cases and recommended a further assessment of angiosarcoma incidence near to incinerators during a more recent period.
30. Members commented with respect to the assessment of angiosarcoma that the uncertainty associated with the derivation of the expected number of angiosarcoma cases was not apparent from the way in which the data had been presented in the report. Giving estimated relative risks with confidence intervals (CI) gave a spurious impression of precision even though the large confidence intervals reflected the variability of the data. Members also commented that further explanation on page 12, paragraph 1, where a ratio of 4 controls per case might be expected whereas a ratio of 6.6:1 was observed, was needed in order to explain the approach used. Members also agreed that the data in Table 4 on page 22 could be better presented by carrying out a test for heterogeneity, and a trend test, further annotations to the table would also make it more clear.
31. Members agreed that overall the data suggested that residence for >10 years near to municipal solid waste incinerators was associated with a very small number of excess primary liver cancer but that it was still possible that the effect could be explained by confounding.
32. Members agreed that a statement could be drafted outlining the reasons for the SAHSU studies, the methods adopted and results obtained and the evaluation of the significance of the results. Members did not feel that a further study looking at a more recent time period was warranted based on the fact that many municipal incinerators had now been closed and that emission of pollutants was very low from those that remained.
ITEM 9 PAPERS FOR INFORMATION
9.1 Trends in breast cancer incidence and mortality CC/98/26
33. Members noted the problems of deriving useful information from reporting of incidence rates by trend. Much could simply be explained by mammographic screening, diagnostic criteria etc. Members also discussed alcohol consumption and questioned why there appeared to be an increase during 1988-9. However it was noted that these were only crude estimates of exposure based on alcohol sales per capita.
9.2 COC views on transgenic tests (published paper) CC/98/27
34. This paper was circulated for information only.
9.3 Additional papers on transgenic tests CC/98/28
35. This paper was circulated for information only. Members commented that the Committee should reconsider the use of short-term tests in transgenic animals during 1999.
ITEM 10 ANY OTHER BUSINESS
36. Members were told that the Cabinet Office had issued a model code for members of Advisory Non-Departmental Public Bodies. The secretariat would be reviewing the document with respect to applying the guidance to COT/COC/COM.
10.2 Food Standards Agency (FSA)
37. Members were told that the setting up of the FSA had been delayed as primary legislation was needed, and it was therefore unlikely that the FSA would be established before late 2000 or early 2001.
38. The meeting ended at 2.27 pm. The secretariat wished all members a happy Christmas and happy New Year.
ITEM 11 DATE OF NEXT MEETING: 18 MARCH 1999
ACTIONS
| Item no | Action | Responsibility |
| 4 | Re-draft conclusions on chlorinated drinking water and cancer | Secretariat |
| 6 | Re-draft statement on organochlorine insecticides and breast cancer | Secretariat |
| 7 | Outline proposal for systematic scoring of epidemiological studies and a meta-analysis | Secretariat |
| 7 | Draft a holding statement for the White Paper on alcohol and breast cancer | Secretariat |
| 8 | Re-draft conclusions on SAHSU study on municipal waste incinerators and liver cancer | Secretariat |