Annex H
Statement by the Committee on the Carcinogenicity of Chemicals in Food, Consumer Products and the Environment to SCOTH on Environmental Tobacco Smoke (ETS) and Lung Cancer
Introduction
1. We have been asked by the Scientific Committee on Tobacco and Health (SCOTH) to review a submission from the Tobacco Manufacturers Association (TMA) comprising a meta-analysis of epidemiological data and supporting references, and a separate meta-analysis paper prepared by Dr A Hackshaw and Professor N Wald (a member of SCOTH). The data provided by the TMA comprised three volumes of reviews and references originally received by the SCOTH secretariat in 1994, and updated in February 1995. A key part of the TMA submission was a meta-analysis of epidemiological studies prepared by Mr P N Lee which was updated in December 1996. The TMA recently submitted 3 additional supplements dated January 1997 dealing with; misclassification bias, dose response (with and without exposed groups), and use of cotinine as a biomarker for exposure to ETS*. We considered all of the submitted information at two meetings in 1997. A further meta-analysis report prepared by an ad-hoc European Working Group was also considered.1 We have also considered additional published literature on the formation and composition of ETS, the results obtained in animal experiments involving exposure to surrogates of ETS, and information regarding investigations to evaluate the potential genotoxicity and biological interactions of ETS in humans published up to June 1997.
2. Smoking tobacco is the predominant cause of lung cancer with approximately 90% of lung cancer deaths in Western populations attributable to cigarette usage.2,3,4 A lower percentage of lung cancer deaths may be attributed to tobacco smoking in developing non-Westernised populations.4 A number of epidemiological assessments undertaken by national regulatory agencies have reported a small but statistically significantly elevated relative risk for lung cancer in passive smokers of between 1.1 to 1.3,3,5,6 whereas other reviewers7-11 concluded that the observed association is due to uncontrolled confounding and biases in these analyses. However, since many individuals within the population are exposed to ETS, it is important to resolve the scientific issues particularly as only a small increase in risk would be associated with many hundreds of deaths due to lung cancer per year.
3. Regarding the structure of our review, it was agreed to consider firstly the nature and composition of ETS followed by information on exposure and uptake of genotoxic components (eg adduct studies) with particular reference to the lung as the target organ. Finally to critically review the submitted epidemiological meta-analyses. All of the available information has been evaluated in accordance with our guidelines12and also with regard to the criteria proposed by Sir Austin Bradford-Hill.13
These latter criteria, which are listed below, are generally regarded as being valuable in the consideration as to whether or not an association between an outcome (in this case lung cancer) and a putative risk factor (passive smoking) is causal.14 A specific reference to each of these criteria in respect of passive smoking and lung cancer has been included in our discussion.
Bradford-Hill criteria
Strength
Consistency
Specificity
Temporality
Biological gradient
Plausibility
Coherence
Experiment
Analogy
Composition of ETS
4. An essential part of our evaluation concerned the chemical composition of ETS and a comparison of this information with data on the composition of mainstream smoke (MS). There is extensive literature on the presence of chemicals in smoke from cigarettes and other tobacco products and many reviews of this information are available.3,4,15-19 ETS consists predominantly of aged diluted sidestream smoke (SS) and some exhaled MS with each type of smoke comprising both a particulate and vapour phase. MS is derived from direct inhalation of smoke from the mouth end of a cigarette whilst SS is the material released directly into the air from the burning tip of the cigarette plus that which diffuses through the cigarette paper. The physical and chemical characteristics of ETS are dynamic and differ significantly from MS and fresh SS. The size of ETS particles decreases rapidly with time due to evaporation of volatile constituents and thus ETS particles are usually smaller than MS particles (ETS particles are approximately 0.1-0.25 µm MMAD whereas MS are approximately 0.1-0.9 µm MMAD).3,5,15,17,20 The chemical composition of ETS also changes rapidly with aging and dilution.21 Nicotine, which is tobacco specific, is present predominantly in the vapour phase of ETS (ca 95%) with a relatively small amount in the particulate phase (ca 5%).3 Concentrations of ETS particulate nicotine rapidly reduce due to evaporation from particles whilst the concentration of nicotine in vapour may reduce due to adsorption onto surfaces.315
5. MS has been the subject of extensive investigation and approximately 4,000 chemicals have been identified to date comprising about 95% of the MS weight.3,4,5 About 10% of these chemicals have been quantified in both MS and SS and these include a lengthy list of known human carcinogens such as 2-naphthylamine, 4-aminobiphenyl, arsenic, hexavalent chromium, vinyl chloride, benzene and a number of genotoxic animal carcinogens that are regarded as potential human carcinogens such as certain polycyclic aromatic hydrocarbons (PAHs, eg benzo(a)pyrene) and nitrosamines (including the Tobacco Specific Nitrosamines (TSNAs) NNK and NNN).3,4,5,15 Yields per cigarette of some carcinogens have been reported to be greater in SS compared to MS22-28 as shown in the table below which presents some selected data from the United States National Research Council (NRC) review.5 Yields of some individual chemicals including a number of carcinogens present in SS have been reported to be relatively constant between different commercial brands including filter and non filter brands of cigarettes. 22,24,27
6. One Research group documented evidence that the use of filters reduced MS emissions from cigarettes but had little effect on SS emission of a number of carcinogens22 Thus some reviewers consider that it is misleading to place too much emphasis on MS/SS ratios.3,15 However we consider it important to note that the data suggest that all three types of smoke MS, SS and ETS contain the same carcinogens and although there will be quantitative differences in composition between different types of smoke, it is likely that the exposure of active and passive smokers to carcinogens will be qualitatively similar. A critical review of the available exposure data on ETS with particular consideration of derived doses of carcinogens in the lung (the target tissue) is given below.
Exposure to carcinogens present in ETS
7. We have considered the available exposure data with particular consideration of the potential exposure of the lung to ETS particles and carcinogens adsorbed to these particles. Several reviews of ETS exposure studies are available.3,5,15,17 The majority of these studies have involved either static or personal monitoring of exposure to carbon dioxide, nicotine, total or respirable particles or ETS particles (estimated by UV or florescence light techniques, or as solanesol particulate material; solanesol is a tobacco leaf constituent).3,5,15 There are a number of recent examples of both static monitoring studies29-33 and personal monitoring studies.34-36 Fewer investigations have reported data on actual exposures to carcinogens present in ETS in field studies (ie under prevailing ambient conditions without manipulating either smoking or environmental conditions).3,5,37 However, there are data to show increased concentrations of carcinogens in indoor air either during or following smoking in respect of benzene, polycyclic aromatic hydrocarbons and nitrosamines thus providing some data on exposure to carcinogens from ETS in field studies.37-41 Some reviewers have commented on the poor control for extraneous non tobacco related sources of carcinogens in the available field studies of indoor air.15,37 Many of the carcinogens which can be found in indoor air such as benzene, polycyclic aromatic hydrocarbons and some volatile nitrosamines can be derived from several sources other than ETS.15,43,44 Exposure to these chemicals will vary depending on location (ie at home, work, or at public venues, during transportation or resulting from leisure activities), local environmental conditions such as cooking of foods and ventilation, and air pollution. A limited number of exposure studies have reported increased concentrations of TSNAs in ETS,15 or in SS.45 One report has documented increased air concentrations of TSNAs (NNK and NNN) derived from ETS in a variety of situations including restaurants, bars and trains.40
8. Quantifying exposure to the carcinogens in ETS and in particular dose levels in the lung is complicated particularly as the chemical composition of ETS rapidly changes depending on factors controlling the levels of MS and SS such as the number of smokers present, the building or room occupation density, size of building/room, number of cigarette or other tobacco products smoked over a given period, individual smoking patterns (puff rate, inhalation volume and duration) and factors controlling losses such as degradation/modification of vapour and particulate ETS constituents through chemical reaction or UV light, and the dilution of ETS constituents due to ventilation, mixing of components (ie homogeneity of ETS) and/or absorption and desorption from surfaces in the room.3,5,15,17,46-48 To illustrate the high potential for variation in air levels of ETS, the United States National Research Council (NRC) modeled air levels of respirable particles (RSPs, <2.5 µm) for a range of conditions expected to be encountered in private residences with one smoker consuming 1-2 cigarettes per hour and found RSP levels varied by two orders of magnitude from approximately 17-5,000 µg/m3.5
9. A large number of the carcinogens associated with ETS are present in the particulate phase. The fraction of ETS particles deposited in the respiratory tract during passive smoking was reported to be 11 ± 4%, ie lower than the fraction of MS particles deposited in the respiratory tract of active smokers (47% ± 13%).49,50 Data from the ICRP66 Lung model reported in the Department of Health Committee on Medical Effects of Air Pollutants report on non-biological particles and health suggest that approximately 42% of 0.05 µm particles and 29% of 0.2 µm particles are deposited in the respiratory tract with a significant proportion of these particles reaching the alveoli.51 These data suggest that ETS particles (ca 0.1-0.25 µm) will penetrate to all regions of the respiratory tract. One group of investigators has calculated that a higher deposition of ETS particles compared to MS particles will occur in the terminal bronchioles and alveoli of the lung.52 The respiratory epithelium of the human lung contains cells with appropriate metabolising capacity to activate carcinogens associated with ETS particles (for example polycyclic aromatic hydrocarbons,53 and tobacco specific nitrosamines such as NNK54).
10. Overall we consider that there are sufficient data to conclude that passive smoking results in an increased dose of genotoxic carcinogens to the respiratory tract including the alveolar region of the lung. In the following section we review the available studies which have investigated the biological properties of ETS particles.
Biological properties of ETS
11. ETS particles contain adsorbed genotoxic carcinogens. The following section presents a review of the biological properties of ETS particles and in particular an assessment of the mutagenic potential of urine samples obtained under field conditions and an evaluation of studies in animals and individuals exposed to ETS. In considering the available studies of the biological properties of ETS, we have paid particular attention to information which is important in assessing whether passive smoking results in exposure to and activation of genotoxic carcinogens in the lung. We have compared exposure data reported in these investigations with published information from field studies15 in order to evaluate degree of exposure to ETS, although we note that only a limited assessment of exposure is possible.
Mutagenic chemicals adsorbed to particles
12. Several research groups have used air sampling techniques in field studies to collect ETS particles and similar methods to collect SS particles during exposure studies. Solvent extracts made from these particles tested in bacterial mutagenicity tests showed the presence of adsorbed mutagenic chemicals which were active in both the presence and absence of an exogenous metabolising fraction.39,55-62 It is difficult to compare the results of the field studies in view of differing methods used in these investigations, the results of which depend heavily on the rate of smoking, sampling methods, number of particles collected by filters, solvent extraction methods, the mutagenicity test methods adopted and the possible influence of confounding sources of air particles containing adsorbed mutagenic substances. Although data regarding objective measures of actual exposures to ETS in these studies were incomplete, we conclude that the weight of evidence supports the view that exposure to mutagenic particles present in ETS occurs under a wide range of field conditions and therefore it is likely to occur under all conditions of passive smoking.39,57,61
Studies in animals
13. Exposure of mice to very high levels of fresh SS is clastogenic inducing micronuclei in polychromatic erythrocytes and exposure of rats to very high levels of either aged diluted or fresh SS induces DNA adducts in a variety of tissues such as the heart, lung, larynx and bladder.63-66 It has been established that MS is carcinogenic in hamsters and rabbits exposed by inhalation or following the application of MS condensates to the skin of mice and rabbits or intrapulmonary injection in rats.4 MS condensates may also act as tumour initiators and promoters in animals.4 SS is carcinogenic in rats when implanted into the lung67 or in mice following skin application.68,69 The results of the skin painting studies in mice have also suggested that on a gravimetric basis the carcinogenic potential of SS condensate exceeds that of MS condensate.68,69 These data show that whole MS and its condensate and SS condensates are carcinogenic in animals and hence we consider it is likely that ETS will also be carcinogenic to animals. However, we note that there are no appropriate life-time bioassays using ETS available to confirm this. We consider that the recent inhalation study where a carcinogenic response was documented in strain A mice exposed to extremely high levels of SS reinforced with some MS was of very limited value and cannot be used to predict hazards to humans.70,71 Evidence of reversible hyperplasia and squamous metaplasia in the nasoturbinates accompanied by active chronic inflammation have been documented in short term inhalation studies of aged diluted SS,72,73 but the relevance of these findings to the potential carcinogenicity of ETS is unclear.
Studies in humans
14. The biological effects of exposure to ETS have been examined in studies involving the measurement of metabolites of carcinogens and the presence of mutagenic substances in urine from exposed individuals. Other relevant studies have investigated chromosomal aberrations and markers of DNA damage (SCEs) in blood lymphocytes and the detection and quantification of carcinogen adducts with DNA and proteins such as haemoglobin or albumin.33,45,74-85 It was reported in the previous section of this statement (see paragraphs 7-10) that exposure to ETS occurs by inhalation and ETS particles are deposited throughout the respiratory tract which has the necessary metabolic capability to activate carcinogens present in ETS. We therefore consider that the presence of carcinogens and/or their adducts in blood or urine provides clear evidence of exposure of the lung to the ultimate genotoxic carcinogens.
15. There is evidence of a small increase in the concentration of mutagenic substances in urine samples taken from passive smokers in a number of investigations where small groups of individuals were exposed to high levels of ETS for periods of 5-8 hours.78,79,83Only one of these studies included partial control for dietary confounding which has been reported to affect the excretion of mutagens in the urine of active smokers.86 A further exposure study where a small group of subjects were maintained on controlled diets did not find a significant increase in the excretion of urinary mutagens following exposure to high levels of ETS for 8 hours.33,82 Limited evidence of increased urinary excretion of mutagenic substances following exposure to ETS has been documented in a small survey of waiters and waitresses80 and in a small survey of blood donors.87 No evidence for an increase in chromosome aberrations in peripheral lymphocytes was documented in one study involving waiters exposed to ETS in restaurants74 or in a number of investigations which considered sister chromatid exchanges in blood lymphocytes74,79-81,84
16. Some more recent studies examined carcinogen DNA or protein adducts in passive smokers.33,75-77,81,85 No increase in 32P-postlabelling of DNA was noted in blood monocytes taken from volunteers exposed to high levels of ETS for 8 hours.33 However, we considered that sampling of blood monocytes was not the most appropriate technique for monitoring exposure to tobacco smoke carcinogens, even in heavy smokers.88-91 Protein adducts can serve as surrogates for DNA adducts, particularly at low exposure doses92-94 and thus most recent attention has therefore been focused on the measurements of protein adducts. A short resume of the main results from three critical studies is presented below.
17. Crawford et al found a statistically significant increase in protein adducts of polycyclic aromatic hydrocarbons using albumin as a marker in children whose mothers smoked compared to children whose mothers did not smoke75 We note that elevated plasma cotinine was also found in children whose mothers smoked and consider that this study was adequately performed. MaClure et al measured adducts of 4-aminobiphenyl (4-ABP)and 3-aminobiphenyl (3-ABP) with haemoglobin following hydrolysis to release these aromatic amines. For 4-ABP adducts there was substantial variability in the results limiting the conclusions that could be drawn. Adducts of 3-ABP were more significantly associated with passive smoking.76 Hammond et al used the same assay as MaClure et al to examine the levels of 4-ABP haemoglobin adducts in pregnant women. Among non-smokers the levels of 4-ABP adducts increased with exposure to ETS. We have considered the results of this study77 and the subsequent correspondence relating to it95-96 and consider that the investigation was adequately conducted and results obtained were valid. It has been demonstrated that 4-ABP exposure in individuals with no history of occupational exposure to this chemical is predominantly derived from tobacco smoking94 and thus the results obtained by Hammond et al provide good evidence that low level exposure to ETS can result in the absorption of genotoxic carcinogens. In a separate study Hecht et al found increased excretion of urinary 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), a specific marker for exposure to the tobacco specific carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) in a small group of 5 individuals exposed to a high level of fresh SS smoke for 3 hours.45 This study provides good evidence to support the view that passive smoking results in exposure of the respiratory tract to tobacco specific carcinogens.
18. Exposure to ETS over a wide range of exposure levels, including those normally encountered in homes, at work and in public places can lead to the inhalation and delivery of genotoxic carcinogens to all parts of the respiratory tract. Furthermore such compounds will be in contact with cells capable of metabolic activation to produce the proximate carcinogens. These data give rise to concern regarding an increased risk of lung cancer in passive smokers. The available information on passive smokers is consistent with that reported for current cigarette smokers where elevated levels of DNA adducts have been documented in samples of lung tissue.88,97-102 The COC advice on genotoxic carcinogens is to make the prudent assumption that any exposure may be associated with some increased health detriment.12 This policy is further supported in this specific instance by the approximately linear dose-response relationship between daily consumption of cigarettes by active smokers and lung cancer risk14 which we consider is consistent with a lack of a threshold. Thus exposure to ETS may be associated with an increased risk of lung cancer, but it is not possible on the basis of these data to make any estimate of the putative increased risk.
|
