CONSIDERATIONS OF THE CHEMICAL COMPLEXITY OF ETS WITH REGARD TO INHALATION STUDIES C.J. Proctor and G. Smith BAT(UK&E) R&D Centre, Regents Park Road, Southampton, S09 I PE, U.K. SUMMARY Any inhalation study that investigates Environmental Tobacco Smoke, ETS, is nsidering a very complex entity. ETS contains numerous chemicals that are continuously changing both in their absolute concentration, in the ratio of concentration between one and another, and even in their particulate to vapour phase distribution. Moreover, when considering ETS in real-life situations, many of the chemical components of ETS will be present as a result of sources other than tobacco smoking. This paper emphasises that the chemical and physical nature of ETS must be considered,in the design and interpretation of any inhalation study on ETS, and illustrates the difficulty in defining precisely what constitutes ETS in such studies. INTRODUCTION Environmental Tobacco Smoke, ETS, is a complex mixture of chemicals found in air as a result of tobacco smoking. Some reports have claimed that exposure to ETS is harmful to the health of the non-smoker (1, 2). This issue has been discussed by scientists and the medical profession for over a decade, and although knowledge has increased over this period, it is still the subject of considerable scientific controversy. The claimshave primarily focused on the suggestion that non-smokers exposed to ETS tpe a sl;,,htly elevated risk of developing lung cancer than non-smokers not so exposed. This conc!usion has mainly been based on combining the results of epidemiological studies that have all been stated to be, when taken individually, inadequate in that they did not measure and control for all potentially confounding variables (1). Moreover, the claimed association is very small, and several experts in epidemio~oqy have suggested thatt, on the existing data, it is impossible LC, draw firm conclusions as to whether ETS exposure is statistically associated with the development of lung cancer in nor.-smokers (3, 4), as small relative risks may well be below the resolution ol the epiderniological techniques (5). CD BATCo document for Legal Services: Health Canada 18 October 1999 Parallel to these efforts many researchers have been investigating dosimetry. and to a lesser extent toxicology, in relation to ETS. Although intended to investigate ETS, these studies include observations an fresh sidestrearn smoke, highly concentrated ETS, as well as realistic levels of ETS. The nature of ETS makes all such studies difficult. ETS is both chemically highly complex and unstable in that it experiences continual temporal and spatial variations. tvioreover, ETS will never be found in isolation in a real-world situation, but rather in combination with chemicals arising from many other sources. The aim of this paper is to illustrate the point that any epidemiology, closimetry or 1,0xicology study should, in their design and interpretation, give due consideration to 6sie complex and unstable nature of ETS. % Ar, THE CHEMICAL AND PHYSICAL NATURE OF ETS ETS,originates mainly from the sidestream. smoke of a burning cigarette, though the mainstream smoke exhaled by the smoker will also contribute. Sidestream smoke, which is the smoke that is convected away from the lit end of the cigarefte, has been investigated by several researchers and has been shown to consist of numerous chemicals. The constituent present in the greatest amount is carbon dioxide (this making up aroun'.4 65% of the total weight). Nicotine is generally the most abundant of the volatile organic chemicals. Broadly the same chemicals present in mainstream smoke are also present in sidestrearn smoke, though their relative yield per cigarette (the mainstream to sidestearn smoke ratio) is highly dependent on the compound considered. Many of the chemicaisidentified in fresh sidestream smoke arp. only found at trace concentrations. Typically. the fresh sidestream smoke of one cigarette would contain around 4 milligrams of nicotine; many of the other constituents would be found at nanogram per cigarette levels. Cigarettes, cigars andpipes will all give quite different sidestream yields, both in quantity and in composition. Moreover, when considering cigarettes, length and weight of .obacco burnt, and the way in which the smoker uses the cigarette, will all affect the sidestreEn yield. The quantity and composition of exhaled mainstream smoke may also varlbetween smokers and the product being smoked. Some work indicates that exhaled s.-noke is a significant contributor to the total ETS particulates, though this is 4~::. CD C:) QtN r\j ON BATCo document for Legal Services: Health Canada 18 October 1999 not the case for volatile chemicals, and so the quantity of exhaled smoke will influence the composition and overall levels of ETS. Thus the origin of ETS is subject to some considerable variability. Once the sidestrearn and exhaled mainstream smoke has been released into the atmosphere, it experiences a large dilution with the ambient air. This has several important implications. The first is that the concentrations of the individual chemicals are dramatically reduced; nicotine concentrations in an environment containing ETS would typically be found at around 5 gg/m3, and the majority of the chemicals identified in fresh sidestrearn smoke are virtually immeasurable in ETS. The second is that the particles undergo considerable evaporation. Radioactive tracer studies have indicated that this evaporation takes place within a few minutes after the forTation of the smoke, and other studies suggest that 20 to 30% of the original matter of the particles is lost by evaporation during this early stage of ageing. The result of this is that many compounds, such as nicotine, associated with the particulate phase of mainstream and fresh sidestrearn smoke, are found in the vapour phase in ETS. These findings have asignificant impacton the manner by which ETSsubsequently decays, and presumably on the uptake by the body of various compounds. The distribution of chemicals between particulate and vapour phase will be dependent both on the volatility of the individual chemical, the age of the smoke and on the concentration of the smoke. Hence studies that use exposures of fresh, highly concentrated smoke will be dealing with something quite different from the aged, dilute ETS experienced in the reai world. % a f Ina x1mum BUILD-UP AND DECAY PROFILES observed level (10ml room, 9 air changesihour, 10% ventilation, 200C, 55% RH, 2 cigarettes Smoked) 80 - co Nicotine Total Hydrocarbons Respirable suspended particulates 60 40 i % I , 20 ;Z) so 90 120 Time (.n:mnas) FIGURE 1 CD CD BATCo document for Legal Services : Health Canada 18 October 1999 The decay of ETS from the atmosphere will depend on many factors, though the most important is the ventilation conditions applied to the room. Moreover, different chemicals will decay at diff erent rates. Gaseous compounds such as carbon monoxide and carbon dioxide decay relatively slowly and are highly dependent upon ventilation rates. whilst nicotine disappears from an indoor environment rapidly and is also likely to be affected by other factors such as the furnishings. The relative decay rates of carbon monoxide, respirable suspended particles, total hydrocarbons and nicotine are illustrated for a single ventilation, temperature and humidity condition in Figure 1. Therefore, in a situation where all of the smokers in a room started smoking the same product at the same time, then at any subsequent point in time the ETS in that room will be different both in absolute concentration and in the relative ~Poncentrations of the various components. This variation is compounded in real-life by different products being smoked at different times. ETS IN THE REAL WORLD No indoor air environment contains ETS in isolation, but rather in combination with chemicals arisingfrom avarietyof different sources (6). These includebuilding materials and furnishings, cooking and heating fuels, aerosol propellants, cleaning agents, and the outdoor air (containing motor vehicle exhausts, etc). Research has shown that although ETS is the most visible portion of the chemicals in indoor air, it is by no means .the largest contributor, and that non-smoking environments are found to have similar chemical burdens to smoking environments (7). ONE S,.'.OKER IN 10TAL 1QN COUNT 100% OFFICE j! i FIGURE 2 Chromatographic profiles of i i offices of multiple occupant similar size, ventilation conditions and total number of TOTAL 10?4 COUNT NO S,.*,GKERS IN occupants. One office was OFFICE occupied by a smoker and three non-smokers, whilst the other was entirely Occupied by non-smokers. 4~. lu 21) 30 10 CD Nj (a~ co BATCo document for Legal Services: Health Canada 18 October 1999 To illustrate this point, Figure 2 compares chromatographic profiles acquired in a smokers'and non-smokers' office in the same modern air-conditioned building. The profiles show the number of stable volatile compounds present in the air of each office, the details of this procedure having been previously published (7). The two offices are of similar size and occupancy. This demonstrates that the air of this environmentis laden with many chemicals, and that ETS adds little, apart from nicotine, to the total burden of volatile chemicals. This supports previous work by other authors. Table 1 summarises data acquired from ten offices in this building, each having been investigated five times (8). Table 1: A Comparison of the arithmetic mean concentrafion of some of the chemicals found in the air of smokers' and non-smokers' offices. Smokers'Offices Non-Smokers'Offices (Arithmetic mean) (Arithmetic mean) Nicotine (jig/e) 6 0.6 Respirable Suspended particulates (pg/m:) 103 90 Carbon monoxide (ppm) 1.4 1.2 Carbon dioxide (pprn) 590 530 Benzene (4g/rn3) 13 12 Styrene (gg/e) 14 17 ffVp - Xylene (pgle) 73 69 (i - Pinene (ggle) 4 4 Undecane (ggte) 5 5 Hence, in the smokers'office 2TS is conr;buting to the levels of some of the chemicals in the air, but ETS is superimposed on a background chemical burden that is much larger than the ETS contribution. This exarnple is not atypical, and non-smckers never exposed to ETS will be exposed everl day ol 'their lives to a cocktail of chemicals arising from various sources. BATCo document for Legal Services : Health Canada 18 October 1999 So, to summarise all these points, ETS is a highly complex entity. It is constantly changing both in its absolute concentration, the relative concentrations of its various components and even in its vapouriparticu late phase distribution. Moreover, in real-world situations ETS is found in combination with chemicals originating from other sources, many of which are common to ETS, and of these multiple sources, ETS is often a minor contributor. THE IMPACT OF THE NATURE OF ETS ON INHALATION STUDIES The complexity of ETS makes the design and interpretation of toxicology and epidemiology studies more difficult than might be encountered with investigations of single substances such as benzene or asbestos. The following discusses some of the potential problems facing toxicity and epidemiological studies of ETS. 1 . Toxicily Studies a) In-vitro : Because ETS is only found in low concentrations in real-world situations, the key problem facing in-vitro studies is the collection of sufficient material to evaluate in -the test system. This generally means that samples are obtained by drawing air through a collection system (often a particulate filter~ at relatively high fIcAv rates for many hours. Table 1 indicates that typically ETS contributes around 13 ~Lg/M3 Of respirable suspended particulates to the environment studied (RSP in smokers'off ice minus RSP in non-smokers' off ices~. Hence, to collect just one milligram of material, some 77 m' of air would have to be drawn through the filter (requiring almost 8 hours continious sampling at 10 m3lhr). To allow this there Must be a continious generation of ETS. It'is obviously important that the sample is representative of ETS, and so account must be taken of the content of volatile and semi-volatile stripped off the collection system during the sampling period. If this is not done, then it is invalid to directly relate results from ETS samples to samples of mainstream smoke (which would be collected over a short period of time, at low flow rate, and would not be stripped of semi-volatile chemicals). Again the office study indicates a considerable back~round of other material in the air, and so great care should also be taken to acquire appropriate controli'background samples. b. CD CD BATCo document for Legal Services: Health Canada 18 October 1999 b) In-vivo :Animal exposure chambers designed to evaluate ETS often generated sidestream smoke externally, which is subsequently delivered to the w1imals. As previously described, both the age and the concentration of the smoke have significant implications upon the refative levels of the chemicals present and upon the particulateivapour phase distribution. "Fresh" or unrealistically high concentrations of ETS will distort this distribution in favour of the particulate phase, hence affecting the site in the body of the animal that individual chemicals will eventually reach. Moreover, Ingebrethsen (9) has shown that the mean particle diameter of ETS is very dependent upon concentration. As concentration increases, both average and mass median diameters increase somewhat, and the number of particles less than 0.01 p diameter decreases dramatically. This is important in terms of the deposition pattern. Even in sub-chronic exposure studies, there is likely to be a distortion in the relative chemical content of the smoke. AdIkofer et al recently reported the results of such a study (10). Their system delivered concentrations of 4 Mg/M3 of respirable suspended particulates and 25 ppm of carbon monoxide to the animals. From the data in Table 1, which gives typical ETS contributions to particulates of 13 Rg/M3 and 0.2 ppm carbon monoxide (calculated by the difference of the average levels found in smokers' and non-smokers'offices), it would be expected that, with no distortion, 25 ppm CO would be assoc-lated with around 1.6 mg/m3 respirable particulates. Hence there are indications that the ETS has been distorted atthese hicher concentrations; it is unknown whether this has any impact on the results of such studies, but the possibility should be considered. A further factor that should be recognised is that some components of ETS deposit rapidly onto surfaces. If the smoke generation system. is connected to the exposure charnberthrough a narrow tube, then there is the potential for a significant distortion in the chernic~Ll composition of the smoke. 2~ Dosimetr As no 'active ingredient', even if there were to be such a thing, has beer. identified, dosimetri must rely on attempting to find chemical markers that are representative of ETS. Nicotine, and its metabolite cotinine, have most frequently been used as a marker. Although nicotinelcotinine is probably the best biomarker available to date, with both good se';ectivity and sensitivity, it has been shown 'that nicotine decays from an atmospi"are far more rapidly than many other components of ETS and is therefore unrepre-s-antative of many of the other components. Nicotine/cOtinine measurements are ani;7,iicator of whether or not a non-smoking. su'z~act has been exposed to ETS, but are c.,,!y useful in quantifying exposure to nicotina, CZ) BATCo document for Legal Services : Health Canada 18 October 1999 Of other possible biomarkers, carboxyhaemaglobin methods have moderate sensitivity but poor selectivity for ETS (with carbon monoxide arising from other sources). Thiocyanate, hydroxyproline, n-nitrosoproline, thioether and urine mutagenicity measurements all suffer from severe confounders such as dietary effects. 3. EpidernigLogy The possible bias in epidemiologic studies of indoor air pollution due to the misclassification of personal exposures has already been emphasised by several authors. The topic is relevant to ETS because it has been shown that indoor air contains many chemicals whether or not smoking is occurring. Thus any study design should ensure adequate control over differences in the chemical composition of air arising from sources other than tobacco (such as radon, automobile exhausts,, cooking, heating etc). Different lifestyles and socioeconomic groupings will most certainly lead to differences in exposures to chemicals both indoors and outdoors. A rural situation versus an urban situation could be quite significantly different because of the influence of outdoor air. Furthermore, without adequate dosimetry or analytical measurement of environments, estimations of exposure to US may be sus I .pect. Several studies have nov; shown that questionnaires are an extremely poor method of assessing exposure to ETS. Often ETS exposure of a non-smoking spouse has been determined by noting khe number of cigarettes smoked per day by the smoker. Even it the smoker smoked all of these cigarettes inclose proximity to the spouse, this estimate of exposure is still inFppropriate. The concentration of ETS intheairwill depend upon ventilation, sizeof rooms, furnishing materials etc. Moreover, arecent study has shown that different products,,,vill result in different ETS levels. Such confounders should be assessed in the design and interpretation of epidemiological studies, and, where possible, dosimetry should be attempted.simultaneousiy. CONCLUSION This paper describes the complexities of the nature of ETS, and emphasises some chemical and physical considerations that have impact on the design and interpretation toxicology studies. CD r\j BATCo document for Legal Services : Health Canada 18 October 1999 KEY REFERENCES 1. Report of the US Surgeon General, Health Consequences of Involuntary Smoking, US Government Printing Office, Washington D.C., 1986. 2. US National Research Council, Environmental Tobacco Smoke, Measuring Exposures and Assessing Health Risks, National Academy Press, Washington D.C., 1986. 3. N. Mantel, British Medical Journal, 294, 6569, 440-441, 1987. 4. K. Uberla, Int. Arch. Occup. Environ. Health 9. 5, 21-37, 1987. 5. R. Montesano et al. in Biology of Carcinogenesis, Ed. M.J. Waring and B.A.J. Ponder, MTP Press Ltd, 1987, p154 6. US NRC, Indoor Pollutants, National Academy Press, Washington D.C., 198 7. C.J. Proctor, Environ. Tech. Letts 9. 553 - 562, 1988. 8. C.J. Proctor, N.D. Warren and M.A.J. Bevan, Proceedings of the Conference on Present and Future of Indoor Air Quality, Brussels, 14 - 16 February, 1989. 9. B.J. Ingebrethsen, Rec. Ady. Tob. Sci., 9. 54-142,1986. 10. F. AdIkofer et al, in Indoor and Ambient Air Quality, ed R. Perry and P.W. Kirk, Seiper Ltd, London, p 252-258, 1988. A full list of references may be obtained from the author. _r_- CD rQ BATCo document for Legal Services : Health Canada 18 October 1999