Mutariott Research, 279 (1992) 233-238 0 1992 Elsevier Science Publishers B.V. All rights reserved 0165-1218/92/S~S.~)
233
MUTGEN 01771
Passive smoking and sister-chromatid exchanges in lymphocytes Wim J.M.J. Gorgels ‘, Geert van Poppel a, Martin J. Jarvis b, Wilma Stenhuis a and Frans J. Kok a a TN0 Toxicology and Nutrition Institute, Zeist (The Netheriunds) and ’ ICRF Health Behariour Unit, Institute bndm ILLK, I
ofPsychiatry.
(Received 7 October 1991) (Revision received 28 November 1991) (Accepted 28 November 1991)
Keywords: Passive smoking; Sister-chromatid exchanges; Chromosomal damage
Summaty The object of this study was to determine whether exposure to environmental tobacco smoke is associated with DNA damage reflected by the frequency of sister-chromatid exchange (SCE) in lymphocytes. Within a cross-sectional design, 106 male non-smoking adults, employees of two administrative companies, were divided on the basis of self-reported exposure into high and low passive smoking groups. The high exposed subjects (passive smokers, n = 50) lived with smokers, worked with smokers and were exposed to tobacco smoke for an average of 70 h/week. The low exposed non-smokers (n = 56) were exposed for an average of 5 h/week. Plasma cotinine levels for the passive smokers ranged between 0.4 and 9.0 ng/ml (median 1.4 ng/mI), and for the low exposed group between 0.0 and 1.9 ng/mI (median 0.4 ng/ml) (p < 0.~01; Mann-jitney test). No difference was observed between the two groups in the frequency of SCEs in lymphocytes: 4.66 f 0.05 for passive smokers and 4.68 f 0.04 for low exposed non-smokers (mean Lt SEMI (p = 0.80; t-test). Reclassification of subjects on the basis of plasma cotinine levels did not change the results SubstantiaIly. These results are in accordance with observations that the increase in cancer risk due to passive smoking is small in comparison with the increase due to active smoking. The SCE test may be too insensitive to be useful for the evaluation of possible cytogenetic effects related to passive smoking.
Current epidemiological evidence points towards a relation between passive smoking and an increase in lung cancer risk (IARC, 1987; Saracci and Riboli, 1989). This relation seems of small magnitude and the aspect of causality needs to be
Correspondence: Geert van Poppel, MSc, TN0 Toxicology and Nutrition Institute, P.O. Box 360, 3700 AJ Zeist (The Netherlands~,
clarified. Saracci and Riboli (1989) reviewed 11 case-control and 3 cohort studies and reported an overall relative risk of lung cancer related to exposure to environmental tobacco smoke of 1.35 (95% confidence interval 1.20- 1.53). However, many of the studies taken individually found no significant association and the influence of possible sources of bias, mainly in the as~ssment of exposure status, is still being debated (Lee, 1987; Kilpatrick, 1987; Saracci and Riboli, 1989).
The sister-chromatid exchange test in lymphocytes is considered to be an in vivo indicator of DNA damage (Bartsch et al., 1988). In view of recent theories. development of DNA damage is an initial process in chemical carcinogenesis (Iversen. 1988). Although the SCE test has been used for evaluating the genotoxic potential of a variety of mutagenic and carcinogenic agents (Lambert et al., 1982; Carrano, 1982; Das, 1988X its predictive value in reflecting cancer proneness still remains to be established. SCEs in lymphocytes have been reported to be increased in smokers (Husgafvel-Pursiainen et al., 1980, Lambert et al., 1978; Husum et al., 1982; Bala Krishna Murthy, 1979; Wulf et al., 1983) and may thus reflect increased cancer risk related to tobacco smoke inhalation. A few studies reported no difference in the frequency of SCE in non-smoking subjects exposed to environmental tobacco smoke compared to non-exposed non-smokers (Husgafvel-Pursiainen, 1987; Husgafvel-Pursiainen et al., 1987; Sorsa et al., 1989: Collman et al., 19861. However, these studies comprised few subjects, and mainly women, which may have reduced precision as SCEs are reported to vary in women related to hormonal changes in the menstrual cycle (Tucker et al., 1987). We have addressed the topic of whether exposure to environmental tobacco smoke is associated with an increased frequency of SCEs in lymphocytes in a larger group consisting of only male subjects and selected on the basis of detailed self-reported exposure. Plasma cotinine levels were measured as an additional indicator of exposure status. Methods
Study population and data collection Maie adult non-smoking employees of two local administrative companies were invited to participate in this study by an advertisement in an in-house company newsletter. Volunteers were asked to fill in a questionnaire on medical history, vitamin and drug intake, work, hobbies and exposure to tobacco smoke (number of smokers in the home environment, number of smoking colleagues at the work place, and estimated hours of exposure to tobacco smoke weekly). Subjects
were selected from the lower and upper tails of the distribution of self-reported exposure. All selected pa~icipants were healthy, had not smoked actively in the last 5 years and were not exposed to environmental mutagens, radiation or medications that could influence the test. The low exposure group (n = 56) had no smokers in the home environment. They reported a m~imum of 10 h/week passive exposure to environmental tobacco smoke, with the exception of 4 with lo-20 h/week. Only 8 had smoking colleagues at work, and these reported less than 10 h of total passive exposure a week. The passive smoking group (n = 50) all had one or more smokers in the home environment. Eight reported passive exposure of 20-40 h/week, and all others reported at least 40 h of exposure per week. Except for 10 subjects, all had smoking colleagues at the work place. Measurement procedures were performed in the morning and included sampling of blood by venipuncture and determination of body weight and height. All subjects and samples were coded and all data were obtained in a blinded fashion to minimize info~ation bias. Laboratory analysis Venous blood was drawn into sterile vacutainer tubes containing sodium heparin as anticoagulant, and immediately put in a coolbox or refrigerator (0-4’0. Blood cultures were set up within 2-6 h post venipuncture, after the blood was allowed to acquire room temperature for 30 min. 0.5 ml blood was added to 4.4 ml prewarmed culture medium in T-25 tissue culture flasks (Costar Europe Ltd, Badhoevedorp, The Netherlands). This medium consisted of RPM1 1640 medium (Flow Laboratories, Irvine, U.K.) supplemented with 20% heat-inactivated fetal calf serum (Flow) (30 min at 56”C), 2.5% phytohemagglutinin (PHA-15 Wellcome, Weesp, The Netherlands), penicillin (100 IU/ml~, streptomycin (100 pg/ml) (Flow) and 2 mM L-glutamine (Flow). Cultures received 10 pg/ml of bromodeoxyuridine (Brdu, Sigma Chemical Co., St. Louis, MO) and were incubated in the dark for 68 + 1 h at 37°C in humidified air containing 5% CO,. Colcemid (Fluka AG, Buchs, Switzerland) was added at a final concentration of 0.2 pg/ml for the last
2 h of imrbation. The cells were harvested by low-speed centrifugation, washed with phosphate-buffered saline (PBS, pH 7.4, Flow), treated with h~potonic KC1 (0.075 M) and fixed with glacial acetic acid in methanol (1 : 3) overnight in the refrigerator at 4°C. The next morning slides were prepared. Preparations, aged for 3 days, were stained by the fluorescence plus Giemsa method (Perry and Thomson, 1984) to obtain harlequin chromosomes. individual data are the mean counts of 50 metaphases scored by a single observer. Only nuclei with 46 chromosomes were scored. SCEs were scored as color changes in the longitudinal direction of the chromatid, excluding the centromere. Plasma cotinine levels were determined by gas chromatography with a detection limit of 0.1 ng/ml (Feyerabend and Russell, 19901.
Differences in SCE scores between low exposed non-smokers and passive smokers were tested with the unpaired f-test. Differences in cotinine values were tested with the Mann-Whitney rank sum test. Correlation analyses were performed with the Spearman rank correlation test. Two-sided p-values were used. One low exposed subject was excluded because lymphocytes failed to culture.
Results In Table 1 some characteristics of the passive smokers and Iow exposed non-smokers in this study are presented. The mean age of the passive smokers was 37.5 years; for the low exposed nonsmokers it was 38.8 years. A small difference was observed in Quetelet index, passive smokers appearing to be slightly heavier. The contrast in self-reported exposure to environmental tobacco smoke was considerable: 5.1 h/week for the low exposed group compared to 72.8 h/week for passive smokers. This contrast was reflected in differences in plasma cotinine levels as presented in Fig. 1. For the low exposed non-smokers the median plasma cotinine concentration was 0.4 ng,/mnI (range: 0.0-1.9). For the passive smokers the median was 1.4 ng/ml (range: O-4-9.0) ( p < 0.001; Mann-Whitney test). As a consequence, there was a clear correlation between seif-reported exposure and plasma cotinine levels (all subjects: r, = 0.57, p = 0.02). However, within the subgroups this correlation was no longer present (low exposed non-smokers: r, = 0.20. p > 0.2; passive smokers: rs = 0.22, p > 0.20). In Fig. 2 the mean SCE score per cell is presented for both groups. No difference was observed between low exposed non-smokers (4.68 f 0.04 (SEMI) and the passive smokers (4.66 f
TABLE 1 CHARACTERISTICS
OF THE LOW EXPOSED NON-SMOKERS AND PASSIVE SMOKERS
Variable
Low exposed non-smokers (n = 55) (mean + SEMI
Passive smokers (n = 50) (mean f SEMI
Age (years) Body weight (kg) Body height fcmf Quetelet index (kg/m’) Passive smoking (self-reported) (h/week) total at home at work elsewhere Mean SCE (total/fiO) Plasma cotinine (ng/ml) median (range)
38.8 78.1 180.2 24.0
37.5 81.2 180.2 25.0
+ 1.4 k1.2 kO.7 +0.3
5.1 +0.9
** ***
+ 1.4 &I.4 + 1.0 +0.4
72.8 +4.8
0.1
35.3
3.2 1.8 4.68 rt 0.04 0.4 (O.O-1.9)
28.7 9.0 4.66 f 0.05 1.4 (0.4-9.0)
Differences between low exposed non-smokers (Mann-Whitney).
* ***
and passive smokers: * p = 0.80 (r-test); * * P = 0.08 (r-test); * * * P < 0.0001
m
low
gotp
O-O.5
exposed fn=551
0.6-l
pas smoktng group (n=50)
1.1-1.5
1.6-2 Plasma
2.1-2.5 cotmine
2.6-3
3.1-3.5
>3.5
hg/ml)
Fig. 1. Plasma cotinine leveis in low exposed non-smokers and passive-smokers.
0.05 (SEMI) (p = 0.80; r-test). Similarly when subjects were ranked on the basis of their plasma cotinine levels, no differences were observed in mean SCE score between subjects with high pfasma cotinine level (upper tertile (n = 36); plasma cotinine > 1.0 ngfml; mean SCE 4.62 it 0.06 (SEM)) and low plasma cotinine level (lowest tertile (n = 35); plasma cotinine 5 0.5 ng/ml; mean SCE 4.65 + 0.06 (SEM)) (p = 0.61; t-test). In addition no difference was found comparing the subgroup of low exposed non-smokers with the Iowest plasma cotinine values (n = 28; plasma
m
cotinine IO.4 ng/ml; mean SCE 4.66 f 0.06 GEM)) with the subgroup of passive smokers with the highest plasma cotinine values (n = 26; plasma cotinine 2 1.3 ng/ml, mean SCE 4.57 + 0.07 (SEMI) (p = 0.32; t-test). Discussion
In the present study the level of SCE was measured in l~pho~tes from a group of passive smokers and a group of low exposed non-smokers. Both groups consisted solely of male subjects
low exposfXl groLq3 b=55f
3.9-4.1
4.2-4.4
pas. smokmg ~oup h=501
4.5-4.7 SCE
4.0-5.0
(mean of
50
5.1-5.3
5.4-5.6
cells)
Fig. 2. SCEs in lymphocytes of low exposed non-smokers and passive smokers.
237
working in the same two administrative companies and were of comparable age. Within the study design a great contrast between the groups was created by selection of subjects on self-reported exposure to environmental tobacco smoke. Our results show no difference in the mean level of SCE in lymphocytes between groups with widely contrasting exposure to passive smoking. For both groups the mean Ievel of SCE per l~pho~yte in the present study was somewhat lower than in other studies CSorsa et al., 1989; Collman et al., 1986; Husgafvel-Pursiainen, 1987; Husga~el-Pursiainen et al., 1987). However, interlaboratory variation in SCE tests is a wellknown phenomenon which makes comparisons with other studies difficult (Das, 1988). In addition, our criterion of excluding centromeric exchanges may have resulted in lower SCE scores. Plasma cotinine is a sensitive biomarker of recent daily life exposure among stringently defined non-smokers (Jarvis et al., 1984). Despite high self-reported exposure, the levels of plasma cotinine in the group of passive smokers in our study among office workers were low compared to the levels measured by Husgafvel-Pursiainen et al. (1987) in restaurant personnel (mean plasma cotinine 10 ng/ml). However, these people may have been more heavily exposed to environmental tobacco smoke as a result of their occupation. Furthermore interlaboratory variation may have contributed to this difference as the non-exposed subjects in the Husgafvel-Pursiainen study, although comparable in self-reported exposure and occupation with the low exposed subjects in this study, had a mean cotinine level of 5.2 ngfml, which is higher than the passive smokers reported here. Plasma cotinine levels in the present study were similar to those reported by Jarvis et al. (1984). A substantial overlap in cotinine levels between the two groups was observed suggesting that some subjects had been misclassified according to self-reporteci exposure. But after reclassification of all subjects on the basis of their cotinine levels there were still no differences in SCE levels. Furthermore, combining the two methods of exposure measurement by comparing the low exposed non-smokers with the lowest cotinine values with the passive smokers with the highest
cotinine values did not reveal any difference in mean SCE levels. As our results do not show any trend towards an association of increase in SCE level and passive smoking it does not seem likely that the lack of measured effects on the SCE parameter can be attributed to ~s~lassi~~ation. Moreover, interrun variation cannot have influenced the results at the group level, as allocation to measurement days was nearly equal for both groups. Our results are in accordance with previous smaller studies in less homogeneous populations of non-smokers (Husga~el-Pursiainen, 1987; Husgafvel-Pursiainen et al., 1987; Collman et al., 1986). These studies also failed to demonstrate even a tendency for an association between passive smoking and SCE Ievels. As cigarette smokers are generally reported to have only lo-50% higher SCE frequencies than non-smokers (IARC, 1986) and a few studies among smokers even failed to find a relationship with the SCE parameter (Crossen and Morgan, 1980; Hollander et al., 1978; Ardito et al., 1980). the SCE test is probably too insensitive to detect the possible cytogenetic effect of passive smoking. For comparison, the power of our study would have been enough to detect a 5% difference at cy= 0.05 and p = 0.05 (interindividual variance: 0.12). The results of this study are consistent with previous findings that the increase in cancer risk due to passive smoking is smal1 in comparison with the increase due to active smoking. Large studies are needed to quantitate this unambiguously. Acknowledgements
We thank the employees and companies who participated in this study; Nice de Vogel for the SCE analysis; Hanny Leezer-de Hoog for assistance in data collection and administration. This study was supported by the Minist~ of Health, Welfare and Cultural Affairs. References Ardito, G.. L. Lamberti, E. Ansaldi and P. Ponzetto (1980) Sister-chromatid exchanges in cigarette-smoking human females and their newels,
Mutation
Res.. 78. 209-212.
Bal;t Krishna Murth>. P. ( 1979) Frequency of sister chromatid cxchanpcs in cigarette smokers. Hum. Genet.. 52.343-345. Bart.sch. H.. K. Hemminki and LK. O’Neill (Eds.) (1988) Methods for detecting DNA damaging agents in humans: applications in cancer epidemiology and prevention. 1ARC Sci. Publ. No. X9. International Agency for Research on Cancer. Lyon. Carrano. A.V. (19S2) Sister chromatid exchange as an indicator of human exposure. in: B.A. Bridges. B.E. Butterworth and LB. Weinstein (Eds.). indicators of genotoxic exposure (Banbury Report 13). Cold Spring Harbor Laboratory. Cold Spring Harbor. NY. pp. 307-318. Collman. G.W.. K. Lundgren. D. Shore. C.L. Thompson and G.W. Lucier (19,Yb) Effects of cY-naphthoflavone on levels of sister cbromatid exchanges in lymphocytes from active and passive cigarette smokers: dose-response relationships. Cancer Res.. 46. 6452-6455. Crossen. P.E.. and W.F. Morgan (1980) Sister chromatid cxchangc in cigarette smokefi, Hum. Gcnet.. 53. 425-326. Das. B.C. (19%) Factors that influence formation of sister chromatid exchanges in human lymphocytes. CRC Crit. Rev. Toxicol., 19. I. Fryer&end. C.. and M.A.H. Russell (1990) A rapid gas-liquid chromatographic method for the determination of cotinine and nicotine in biological fluids. J. Pharm. Pharmacol.. 42. 450-452. Hollander. D-H., M.S. Tockman. Y.W. Liang. D.A. Borgaonkar and J.K. Frost (1978) Sister chromatid exchanges in the peripheral blood of cigarette smokers and in lung cancer patients: and the effect of chemotherapy, Hum. Cienet.. 4-l. 165-171. Husgafiel-Pursiainen. K. (1987) Sister-chromatid exchange and cell proliferation in cultured lymphocytes of passively and actively smoking restaurant personnel, Mutation Res.. 190. ‘I I-215. Husgafvel-Pursiainen. K.. J. Miiki-Paakkanen. H. Norppa and M. Sorsa (1980) Smoking and Sister Chromatid Exchange, Hereditas. 92. 247-250. Husgafvel-Pursiainen, K.. M. Sorsa. K. Engstriim and P. Einistii (1987) Passive smoking at work: biochemical and biological measurements of exposure to environmental tobacco smoke. Int. Arch. Occup. Environ. Health. 59, 337345. Husum. B.. H.C. Wulf and E. Niebuhr (1982) Increased sister chromatid exchange frequency in lymphocytes in healthy cigarette smokers. Hereditas. 96, 85-88. IARC (1986) Monographs on the Evaluation of the Carcino-
genie Risk of Chemicals to Humans. Vol. 38: Tobacco smoking. IARC. Lyon. IARC (1987) Environmental Carcinogens. Methods of Analysis and Exposure Measurement, Vol. 9: Passive smoking, IARC. Lyon. Iversen, O.H. (Ed.) (1988) Theories of Carcinogenesis, Hemisphere, Washington, DC. Jarvis. M.J., H. Tunstall-Pedoe. C. Feyerabend, C. Vesey and Y. Saloojee (1984) Biochemical markers of smoke absorption and self reported exposure to passive smoking, .I. Epidemiol. Commun. Health. 38, 335-339. Kilpa’rick Jr., S.J. (1987) Misclassification of environmental tobacco smoke exposure: its potential influence on studies of environmental tobacco smoke and lung cancer, Toxicol. Lett.. 35, 163-168. Lambert, B., A. Lindblad, M. Nordenskjtild and B. Werelius (1978) Increased frequency of sister chromatid exchange in cigarette smokers, Hereditas, 88, 147-149. Lambert, B.. A. Lindblad, K. Holmberg and D. Frdncesconi (1982) The use of sister chromatid exchange to monitor human populations for exposure to toxicologically harmful agents, in: S. Wolff (Ed.). Sister Chromatid Exchange, Wiley. New York, pp. 149-182. Lee, P.N. (1987) Lung cancer and passive smoking: association an artefact due to misclassification of smoking habits?, Toxicol. Lett., 35, 157-162. Perry, P.E., and E.J. Thomson (1984) The methodology of sister chromatid exchanges, in: B.J. Kilbey, M. Legator, W. Nichols and C. Ramel (Eds.). Handbook of Mutagenicity Test Procedures. 2nd edn., Elsevier, Amsterdam, pp. 49% 529. Saracci, R., and E. Riboli (1989) Passive smoking and lung cancer: current evidence and ongoing studies at the fnternational Agency for Research on Cancer, Mutation Res., 222, 117-127. Sorsa, M., K. Husgafvel-Pursiainen, H. Jgrventaus, K. Koskimies, H. Salo and H. Vainio (1989) Cytogenetic effects of tobacco smoke exposure among involuntary smokers, Mutation Res., 222, 11 l-l 16. Tucker, J.D. et al. (1987) Variation in the human lymphocyte sister chromatid exchange frequency as a function of time: results of daily and twice-weekly sampling, Environ. Mol. Mutagen., 10, 69-77. Wulf, H.C., B. Husum and E. Niebuhr (1983) Sister chromatid exchanges in smokers of high-tar cigarettes, low-tar cigarettes. cheroots and pipe tobacco. Hereditas, 98, 22522x.
233
MUTGEN 01771
Passive smoking and sister-chromatid exchanges in lymphocytes Wim J.M.J. Gorgels ‘, Geert van Poppel a, Martin J. Jarvis b, Wilma Stenhuis a and Frans J. Kok a a TN0 Toxicology and Nutrition Institute, Zeist (The Netheriunds) and ’ ICRF Health Behariour Unit, Institute bndm ILLK, I
ofPsychiatry.
(Received 7 October 1991) (Revision received 28 November 1991) (Accepted 28 November 1991)
Keywords: Passive smoking; Sister-chromatid exchanges; Chromosomal damage
Summaty The object of this study was to determine whether exposure to environmental tobacco smoke is associated with DNA damage reflected by the frequency of sister-chromatid exchange (SCE) in lymphocytes. Within a cross-sectional design, 106 male non-smoking adults, employees of two administrative companies, were divided on the basis of self-reported exposure into high and low passive smoking groups. The high exposed subjects (passive smokers, n = 50) lived with smokers, worked with smokers and were exposed to tobacco smoke for an average of 70 h/week. The low exposed non-smokers (n = 56) were exposed for an average of 5 h/week. Plasma cotinine levels for the passive smokers ranged between 0.4 and 9.0 ng/ml (median 1.4 ng/mI), and for the low exposed group between 0.0 and 1.9 ng/mI (median 0.4 ng/ml) (p < 0.~01; Mann-jitney test). No difference was observed between the two groups in the frequency of SCEs in lymphocytes: 4.66 f 0.05 for passive smokers and 4.68 f 0.04 for low exposed non-smokers (mean Lt SEMI (p = 0.80; t-test). Reclassification of subjects on the basis of plasma cotinine levels did not change the results SubstantiaIly. These results are in accordance with observations that the increase in cancer risk due to passive smoking is small in comparison with the increase due to active smoking. The SCE test may be too insensitive to be useful for the evaluation of possible cytogenetic effects related to passive smoking.
Current epidemiological evidence points towards a relation between passive smoking and an increase in lung cancer risk (IARC, 1987; Saracci and Riboli, 1989). This relation seems of small magnitude and the aspect of causality needs to be
Correspondence: Geert van Poppel, MSc, TN0 Toxicology and Nutrition Institute, P.O. Box 360, 3700 AJ Zeist (The Netherlands~,
clarified. Saracci and Riboli (1989) reviewed 11 case-control and 3 cohort studies and reported an overall relative risk of lung cancer related to exposure to environmental tobacco smoke of 1.35 (95% confidence interval 1.20- 1.53). However, many of the studies taken individually found no significant association and the influence of possible sources of bias, mainly in the as~ssment of exposure status, is still being debated (Lee, 1987; Kilpatrick, 1987; Saracci and Riboli, 1989).
The sister-chromatid exchange test in lymphocytes is considered to be an in vivo indicator of DNA damage (Bartsch et al., 1988). In view of recent theories. development of DNA damage is an initial process in chemical carcinogenesis (Iversen. 1988). Although the SCE test has been used for evaluating the genotoxic potential of a variety of mutagenic and carcinogenic agents (Lambert et al., 1982; Carrano, 1982; Das, 1988X its predictive value in reflecting cancer proneness still remains to be established. SCEs in lymphocytes have been reported to be increased in smokers (Husgafvel-Pursiainen et al., 1980, Lambert et al., 1978; Husum et al., 1982; Bala Krishna Murthy, 1979; Wulf et al., 1983) and may thus reflect increased cancer risk related to tobacco smoke inhalation. A few studies reported no difference in the frequency of SCE in non-smoking subjects exposed to environmental tobacco smoke compared to non-exposed non-smokers (Husgafvel-Pursiainen, 1987; Husgafvel-Pursiainen et al., 1987; Sorsa et al., 1989: Collman et al., 19861. However, these studies comprised few subjects, and mainly women, which may have reduced precision as SCEs are reported to vary in women related to hormonal changes in the menstrual cycle (Tucker et al., 1987). We have addressed the topic of whether exposure to environmental tobacco smoke is associated with an increased frequency of SCEs in lymphocytes in a larger group consisting of only male subjects and selected on the basis of detailed self-reported exposure. Plasma cotinine levels were measured as an additional indicator of exposure status. Methods
Study population and data collection Maie adult non-smoking employees of two local administrative companies were invited to participate in this study by an advertisement in an in-house company newsletter. Volunteers were asked to fill in a questionnaire on medical history, vitamin and drug intake, work, hobbies and exposure to tobacco smoke (number of smokers in the home environment, number of smoking colleagues at the work place, and estimated hours of exposure to tobacco smoke weekly). Subjects
were selected from the lower and upper tails of the distribution of self-reported exposure. All selected pa~icipants were healthy, had not smoked actively in the last 5 years and were not exposed to environmental mutagens, radiation or medications that could influence the test. The low exposure group (n = 56) had no smokers in the home environment. They reported a m~imum of 10 h/week passive exposure to environmental tobacco smoke, with the exception of 4 with lo-20 h/week. Only 8 had smoking colleagues at work, and these reported less than 10 h of total passive exposure a week. The passive smoking group (n = 50) all had one or more smokers in the home environment. Eight reported passive exposure of 20-40 h/week, and all others reported at least 40 h of exposure per week. Except for 10 subjects, all had smoking colleagues at the work place. Measurement procedures were performed in the morning and included sampling of blood by venipuncture and determination of body weight and height. All subjects and samples were coded and all data were obtained in a blinded fashion to minimize info~ation bias. Laboratory analysis Venous blood was drawn into sterile vacutainer tubes containing sodium heparin as anticoagulant, and immediately put in a coolbox or refrigerator (0-4’0. Blood cultures were set up within 2-6 h post venipuncture, after the blood was allowed to acquire room temperature for 30 min. 0.5 ml blood was added to 4.4 ml prewarmed culture medium in T-25 tissue culture flasks (Costar Europe Ltd, Badhoevedorp, The Netherlands). This medium consisted of RPM1 1640 medium (Flow Laboratories, Irvine, U.K.) supplemented with 20% heat-inactivated fetal calf serum (Flow) (30 min at 56”C), 2.5% phytohemagglutinin (PHA-15 Wellcome, Weesp, The Netherlands), penicillin (100 IU/ml~, streptomycin (100 pg/ml) (Flow) and 2 mM L-glutamine (Flow). Cultures received 10 pg/ml of bromodeoxyuridine (Brdu, Sigma Chemical Co., St. Louis, MO) and were incubated in the dark for 68 + 1 h at 37°C in humidified air containing 5% CO,. Colcemid (Fluka AG, Buchs, Switzerland) was added at a final concentration of 0.2 pg/ml for the last
2 h of imrbation. The cells were harvested by low-speed centrifugation, washed with phosphate-buffered saline (PBS, pH 7.4, Flow), treated with h~potonic KC1 (0.075 M) and fixed with glacial acetic acid in methanol (1 : 3) overnight in the refrigerator at 4°C. The next morning slides were prepared. Preparations, aged for 3 days, were stained by the fluorescence plus Giemsa method (Perry and Thomson, 1984) to obtain harlequin chromosomes. individual data are the mean counts of 50 metaphases scored by a single observer. Only nuclei with 46 chromosomes were scored. SCEs were scored as color changes in the longitudinal direction of the chromatid, excluding the centromere. Plasma cotinine levels were determined by gas chromatography with a detection limit of 0.1 ng/ml (Feyerabend and Russell, 19901.
Differences in SCE scores between low exposed non-smokers and passive smokers were tested with the unpaired f-test. Differences in cotinine values were tested with the Mann-Whitney rank sum test. Correlation analyses were performed with the Spearman rank correlation test. Two-sided p-values were used. One low exposed subject was excluded because lymphocytes failed to culture.
Results In Table 1 some characteristics of the passive smokers and Iow exposed non-smokers in this study are presented. The mean age of the passive smokers was 37.5 years; for the low exposed nonsmokers it was 38.8 years. A small difference was observed in Quetelet index, passive smokers appearing to be slightly heavier. The contrast in self-reported exposure to environmental tobacco smoke was considerable: 5.1 h/week for the low exposed group compared to 72.8 h/week for passive smokers. This contrast was reflected in differences in plasma cotinine levels as presented in Fig. 1. For the low exposed non-smokers the median plasma cotinine concentration was 0.4 ng,/mnI (range: 0.0-1.9). For the passive smokers the median was 1.4 ng/ml (range: O-4-9.0) ( p < 0.001; Mann-Whitney test). As a consequence, there was a clear correlation between seif-reported exposure and plasma cotinine levels (all subjects: r, = 0.57, p = 0.02). However, within the subgroups this correlation was no longer present (low exposed non-smokers: r, = 0.20. p > 0.2; passive smokers: rs = 0.22, p > 0.20). In Fig. 2 the mean SCE score per cell is presented for both groups. No difference was observed between low exposed non-smokers (4.68 f 0.04 (SEMI) and the passive smokers (4.66 f
TABLE 1 CHARACTERISTICS
OF THE LOW EXPOSED NON-SMOKERS AND PASSIVE SMOKERS
Variable
Low exposed non-smokers (n = 55) (mean + SEMI
Passive smokers (n = 50) (mean f SEMI
Age (years) Body weight (kg) Body height fcmf Quetelet index (kg/m’) Passive smoking (self-reported) (h/week) total at home at work elsewhere Mean SCE (total/fiO) Plasma cotinine (ng/ml) median (range)
38.8 78.1 180.2 24.0
37.5 81.2 180.2 25.0
+ 1.4 k1.2 kO.7 +0.3
5.1 +0.9
** ***
+ 1.4 &I.4 + 1.0 +0.4
72.8 +4.8
0.1
35.3
3.2 1.8 4.68 rt 0.04 0.4 (O.O-1.9)
28.7 9.0 4.66 f 0.05 1.4 (0.4-9.0)
Differences between low exposed non-smokers (Mann-Whitney).
* ***
and passive smokers: * p = 0.80 (r-test); * * P = 0.08 (r-test); * * * P < 0.0001
m
low
gotp
O-O.5
exposed fn=551
0.6-l
pas smoktng group (n=50)
1.1-1.5
1.6-2 Plasma
2.1-2.5 cotmine
2.6-3
3.1-3.5
>3.5
hg/ml)
Fig. 1. Plasma cotinine leveis in low exposed non-smokers and passive-smokers.
0.05 (SEMI) (p = 0.80; r-test). Similarly when subjects were ranked on the basis of their plasma cotinine levels, no differences were observed in mean SCE score between subjects with high pfasma cotinine level (upper tertile (n = 36); plasma cotinine > 1.0 ngfml; mean SCE 4.62 it 0.06 (SEM)) and low plasma cotinine level (lowest tertile (n = 35); plasma cotinine 5 0.5 ng/ml; mean SCE 4.65 + 0.06 (SEM)) (p = 0.61; t-test). In addition no difference was found comparing the subgroup of low exposed non-smokers with the Iowest plasma cotinine values (n = 28; plasma
m
cotinine IO.4 ng/ml; mean SCE 4.66 f 0.06 GEM)) with the subgroup of passive smokers with the highest plasma cotinine values (n = 26; plasma cotinine 2 1.3 ng/ml, mean SCE 4.57 + 0.07 (SEMI) (p = 0.32; t-test). Discussion
In the present study the level of SCE was measured in l~pho~tes from a group of passive smokers and a group of low exposed non-smokers. Both groups consisted solely of male subjects
low exposfXl groLq3 b=55f
3.9-4.1
4.2-4.4
pas. smokmg ~oup h=501
4.5-4.7 SCE
4.0-5.0
(mean of
50
5.1-5.3
5.4-5.6
cells)
Fig. 2. SCEs in lymphocytes of low exposed non-smokers and passive smokers.
237
working in the same two administrative companies and were of comparable age. Within the study design a great contrast between the groups was created by selection of subjects on self-reported exposure to environmental tobacco smoke. Our results show no difference in the mean level of SCE in lymphocytes between groups with widely contrasting exposure to passive smoking. For both groups the mean Ievel of SCE per l~pho~yte in the present study was somewhat lower than in other studies CSorsa et al., 1989; Collman et al., 1986; Husgafvel-Pursiainen, 1987; Husga~el-Pursiainen et al., 1987). However, interlaboratory variation in SCE tests is a wellknown phenomenon which makes comparisons with other studies difficult (Das, 1988). In addition, our criterion of excluding centromeric exchanges may have resulted in lower SCE scores. Plasma cotinine is a sensitive biomarker of recent daily life exposure among stringently defined non-smokers (Jarvis et al., 1984). Despite high self-reported exposure, the levels of plasma cotinine in the group of passive smokers in our study among office workers were low compared to the levels measured by Husgafvel-Pursiainen et al. (1987) in restaurant personnel (mean plasma cotinine 10 ng/ml). However, these people may have been more heavily exposed to environmental tobacco smoke as a result of their occupation. Furthermore interlaboratory variation may have contributed to this difference as the non-exposed subjects in the Husgafvel-Pursiainen study, although comparable in self-reported exposure and occupation with the low exposed subjects in this study, had a mean cotinine level of 5.2 ngfml, which is higher than the passive smokers reported here. Plasma cotinine levels in the present study were similar to those reported by Jarvis et al. (1984). A substantial overlap in cotinine levels between the two groups was observed suggesting that some subjects had been misclassified according to self-reporteci exposure. But after reclassification of all subjects on the basis of their cotinine levels there were still no differences in SCE levels. Furthermore, combining the two methods of exposure measurement by comparing the low exposed non-smokers with the lowest cotinine values with the passive smokers with the highest
cotinine values did not reveal any difference in mean SCE levels. As our results do not show any trend towards an association of increase in SCE level and passive smoking it does not seem likely that the lack of measured effects on the SCE parameter can be attributed to ~s~lassi~~ation. Moreover, interrun variation cannot have influenced the results at the group level, as allocation to measurement days was nearly equal for both groups. Our results are in accordance with previous smaller studies in less homogeneous populations of non-smokers (Husga~el-Pursiainen, 1987; Husgafvel-Pursiainen et al., 1987; Collman et al., 1986). These studies also failed to demonstrate even a tendency for an association between passive smoking and SCE Ievels. As cigarette smokers are generally reported to have only lo-50% higher SCE frequencies than non-smokers (IARC, 1986) and a few studies among smokers even failed to find a relationship with the SCE parameter (Crossen and Morgan, 1980; Hollander et al., 1978; Ardito et al., 1980). the SCE test is probably too insensitive to detect the possible cytogenetic effect of passive smoking. For comparison, the power of our study would have been enough to detect a 5% difference at cy= 0.05 and p = 0.05 (interindividual variance: 0.12). The results of this study are consistent with previous findings that the increase in cancer risk due to passive smoking is smal1 in comparison with the increase due to active smoking. Large studies are needed to quantitate this unambiguously. Acknowledgements
We thank the employees and companies who participated in this study; Nice de Vogel for the SCE analysis; Hanny Leezer-de Hoog for assistance in data collection and administration. This study was supported by the Minist~ of Health, Welfare and Cultural Affairs. References Ardito, G.. L. Lamberti, E. Ansaldi and P. Ponzetto (1980) Sister-chromatid exchanges in cigarette-smoking human females and their newels,
Mutation
Res.. 78. 209-212.
Bal;t Krishna Murth>. P. ( 1979) Frequency of sister chromatid cxchanpcs in cigarette smokers. Hum. Genet.. 52.343-345. Bart.sch. H.. K. Hemminki and LK. O’Neill (Eds.) (1988) Methods for detecting DNA damaging agents in humans: applications in cancer epidemiology and prevention. 1ARC Sci. Publ. No. X9. International Agency for Research on Cancer. Lyon. Carrano. A.V. (19S2) Sister chromatid exchange as an indicator of human exposure. in: B.A. Bridges. B.E. Butterworth and LB. Weinstein (Eds.). indicators of genotoxic exposure (Banbury Report 13). Cold Spring Harbor Laboratory. Cold Spring Harbor. NY. pp. 307-318. Collman. G.W.. K. Lundgren. D. Shore. C.L. Thompson and G.W. Lucier (19,Yb) Effects of cY-naphthoflavone on levels of sister cbromatid exchanges in lymphocytes from active and passive cigarette smokers: dose-response relationships. Cancer Res.. 46. 6452-6455. Crossen. P.E.. and W.F. Morgan (1980) Sister chromatid cxchangc in cigarette smokefi, Hum. Gcnet.. 53. 425-326. Das. B.C. (19%) Factors that influence formation of sister chromatid exchanges in human lymphocytes. CRC Crit. Rev. Toxicol., 19. I. Fryer&end. C.. and M.A.H. Russell (1990) A rapid gas-liquid chromatographic method for the determination of cotinine and nicotine in biological fluids. J. Pharm. Pharmacol.. 42. 450-452. Hollander. D-H., M.S. Tockman. Y.W. Liang. D.A. Borgaonkar and J.K. Frost (1978) Sister chromatid exchanges in the peripheral blood of cigarette smokers and in lung cancer patients: and the effect of chemotherapy, Hum. Cienet.. 4-l. 165-171. Husgafiel-Pursiainen. K. (1987) Sister-chromatid exchange and cell proliferation in cultured lymphocytes of passively and actively smoking restaurant personnel, Mutation Res.. 190. ‘I I-215. Husgafvel-Pursiainen. K.. J. Miiki-Paakkanen. H. Norppa and M. Sorsa (1980) Smoking and Sister Chromatid Exchange, Hereditas. 92. 247-250. Husgafvel-Pursiainen, K.. M. Sorsa. K. Engstriim and P. Einistii (1987) Passive smoking at work: biochemical and biological measurements of exposure to environmental tobacco smoke. Int. Arch. Occup. Environ. Health. 59, 337345. Husum. B.. H.C. Wulf and E. Niebuhr (1982) Increased sister chromatid exchange frequency in lymphocytes in healthy cigarette smokers. Hereditas. 96, 85-88. IARC (1986) Monographs on the Evaluation of the Carcino-
genie Risk of Chemicals to Humans. Vol. 38: Tobacco smoking. IARC. Lyon. IARC (1987) Environmental Carcinogens. Methods of Analysis and Exposure Measurement, Vol. 9: Passive smoking, IARC. Lyon. Iversen, O.H. (Ed.) (1988) Theories of Carcinogenesis, Hemisphere, Washington, DC. Jarvis. M.J., H. Tunstall-Pedoe. C. Feyerabend, C. Vesey and Y. Saloojee (1984) Biochemical markers of smoke absorption and self reported exposure to passive smoking, .I. Epidemiol. Commun. Health. 38, 335-339. Kilpa’rick Jr., S.J. (1987) Misclassification of environmental tobacco smoke exposure: its potential influence on studies of environmental tobacco smoke and lung cancer, Toxicol. Lett.. 35, 163-168. Lambert, B., A. Lindblad, M. Nordenskjtild and B. Werelius (1978) Increased frequency of sister chromatid exchange in cigarette smokers, Hereditas, 88, 147-149. Lambert, B.. A. Lindblad, K. Holmberg and D. Frdncesconi (1982) The use of sister chromatid exchange to monitor human populations for exposure to toxicologically harmful agents, in: S. Wolff (Ed.). Sister Chromatid Exchange, Wiley. New York, pp. 149-182. Lee, P.N. (1987) Lung cancer and passive smoking: association an artefact due to misclassification of smoking habits?, Toxicol. Lett., 35, 157-162. Perry, P.E., and E.J. Thomson (1984) The methodology of sister chromatid exchanges, in: B.J. Kilbey, M. Legator, W. Nichols and C. Ramel (Eds.). Handbook of Mutagenicity Test Procedures. 2nd edn., Elsevier, Amsterdam, pp. 49% 529. Saracci, R., and E. Riboli (1989) Passive smoking and lung cancer: current evidence and ongoing studies at the fnternational Agency for Research on Cancer, Mutation Res., 222, 117-127. Sorsa, M., K. Husgafvel-Pursiainen, H. Jgrventaus, K. Koskimies, H. Salo and H. Vainio (1989) Cytogenetic effects of tobacco smoke exposure among involuntary smokers, Mutation Res., 222, 11 l-l 16. Tucker, J.D. et al. (1987) Variation in the human lymphocyte sister chromatid exchange frequency as a function of time: results of daily and twice-weekly sampling, Environ. Mol. Mutagen., 10, 69-77. Wulf, H.C., B. Husum and E. Niebuhr (1983) Sister chromatid exchanges in smokers of high-tar cigarettes, low-tar cigarettes. cheroots and pipe tobacco. Hereditas, 98, 22522x.
