- Email: [email protected]
Factors contributing to biomarker responses in exposed workers
Mutation Research 428 Ž1999. 197–202 www.elsevier.comrlocatermolmut Community address: www.elsevier.comrlocatermutres
Factors contributing to biomarker responses in exposed workers Diana Anderson
)
BIBRA International, Woodmansterne Road, Carshalton, Surrey SM5 4DS, UK Received 29 November 1998; accepted 11 January 1999
Abstract ‘‘ . . . wThxe proper study of mankind is man’’ ŽPope, circa 1733r1734.. Human monitoring fits this notion and monitoring after exposure to genotoxic agents is now an established discipline. It is possible in many situations to identify humans exposed to potentially toxic materials in the workplace and the environment. Responses are often measured in peripheral lymphocytes because these cells can be acquired by a generally socially and ethically acceptable, minimally invasive route. In the early 1960s, chromosome damage in these cells was one of the first endpoints to be used as a biomarker and benzene was one of the first chemicals investigated. Although a causal relationship between chromosome damage and cancer has not been proven, it has been suggested to have some prognostic significance for future cancer onset. With other genetic biomarkers this is as yet not the case, but there are now many biomarkers for different areas of toxicology. Other well-established genetic biomarkers include the detection of hprt mutations, micronuclei and sister chromatid exchanges. However, for interpretation of responses, the issue of confounding factors must be addressed. As in most human studies, there tends to be a high degree of interindividual variability in response to chemical insults. Some non-exposed control individuals exhibit as high a level of damage as some exposed individuals and some of these have levels of damage as low as many of the controls. Thus, it is only the mean values of the groups that can substantiate an exposure related-problem; the data on an individual basis are still of limited use. While human lymphocytes remain the most popular cell type for monitoring purposes, sperm, buccal, nasal, epithelial and placental cells are also used. Confounding factors affect responses in all cell types. There are endogenous confounding factors such as age, sex, genetic make-up and exogenous confounding factors including lifestyle habits such as smoking, drinking, etc. There are biomarkers of exposure, effect and susceptibility and the last may be influenced by the genotype and polymorphism genes existing in a population. From our own studies, confounding effects will be considered in relation to workers exposed to vinyl chloride and petroleum emissions. The relationship between the biomarkers and various factors which influence them is complex. Sometimes the variables are not completely independent of one another. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Confounding factor; Age; Sex; Smoking; Drinking; Lymphocyte; Vinyl chloride; Petroleum emission
1. Introduction ‘‘The proper study of mankind is man’’ w1x. Human monitoring fits this notion and monitoring after ) Tel.: q44-181-6521000; Fax: q44-181-6617029; E-mail: [email protected]
exposure to genotoxic agents is now an established discipline w2x. Human somatic and germinal mutation frequencies may be increased by exposure to a variety of agents. Genic or chromosomal somatic mutations may contribute to acquired disorders such as cancer and germinal mutations and are likely to contribute
0027-5107r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 4 2 Ž 9 9 . 0 0 0 4 7 - 2
198
D. Andersonr Mutation Research 428 (1999) 197–202
to inherited defects in the offspring of individuals exposed to mutagenic agents. There is thus a continuing need to develop and supply methods to study exposed populations at risk. It is possible in many situations to identify humans exposed to potentially toxic materials in the workplace and the environment. Responses are often measured in peripheral lymphocytes because these cells can be acquired by a generally socially and ethically acceptable, minimally invasive route. Radiation was one of the first agents to be examined in this way in the early 1960s and chromosome damage was one of the first endpoints to be used w3,4x. Chemical damage was similarly measured in the mid-1960s w5x. Although a causal relationship between chromosomal aberrations and cancer has not been proven, a prospective study by the Nordic Group that examined 3182 individuals w6x showed a highly statistically significant linear trend for a positive association Ž p s 0.0009. between chromosome damage in peripheral lymphocytes and subsequent cancer risk. The data were not extensive enough to determine if the same associations existed for sister chromatid exchanges and micronuclei. Bonassi et al. w7x also suggested that chromosomal aberrations in circulating lymphocytes were predictive of future cancer onset in humans. Furthermore, many studies have shown a higher level of cytogenetic damage in tumour cells than in normal cells, and an early association of this nature was made by Forni and Moreo w8x. Methods to detect gene mutations have also been long established Že.g., for the HPRT locus. and increased mutation frequencies have been shown to result from genotoxic exposures. Some of the newer monitoring approaches include determination of cytogenetic damage using fluorescence in situ hybridisation ŽFISH. techniques, metabolic genotypes that might show altered sensitivity, such as GSTMI and NAT2 and DNA strand breakage in the single cell gel DNA electrophoresis ŽComet assay.. DNA microsatellite instabilities, oncoproteins Že.g., ras p21. and tumour suppressor genes Žp53. are also involved in the cancer process, and measurement of these has now been added to the monitoring armoury. While human lymphocytes remain the most popular cell type for monitoring purposes, sperm, buccal, nasal epithelial and placental cells are also being used.
These cells can be acquired with minimal invasion of the individual. Exposure to genotoxic agents can result from natural and environmental factors, non-specific contamination, the occupational environment, or industrial accidents. For interpretation of responses, the issue of confounding factors must be addressed. As in most human studies, there tends to be a high degree of interindividual variability in response to toxic insults. Some non-exposed individuals exhibit as high a level of damage as some exposed individuals and some of those exposed have levels of damage as low as many of the controls. Thus, it still pertains that only the mean values of the groups can substantiate an exposure-related problem; the data on an individual basis are still of limited use. Confounding factors affect responses in all cell types. There are endogenous confounding factors such as age, sex, genetic make-up and exogenous confounding factors include lifestyle habits such as smoking, drinking, etc. There are biomarkers of exposure, effect and susceptibility and the last may be influenced by the genotype and polymorphisms in genes existing in a population. From our own studies, confounding effects will be considered in relation to workers exposed to vinyl chloride and petroleum emissions.
2. Materials and methods 2.1. Sampling Interviews were performed using a questionnaire. Blood samples were collected from exposed and unexposed groups. To avoid a seasonal influence, sampling of exposed and unexposed groups was carried out during the same time period. The methods used for the preparation of samples and the details of the studies described in the present communication have been published elsewhere w9– 11x. 2.2. Statistics For the vinyl chloride study, the results were assumed to conform to a Poisson distribution and the
D. Andersonr Mutation Research 428 (1999) 197–202
199
Table 1 Chromosomal abnormalities in workers employed in polyvinylchloride manufacture ŽB and C cells classified according to Buckton and Pike w18x.
Group 1 Group 2 Group 3 Group 4 Controls on-site Controls off-site
No. of workers
%B cells
% Cu cells
% Cs cells
18 21 5 12 19
6.4 6.9 6.6 5.1 3.6
1.72 1.48 1.20 1.10 0.53
0.44 0.30 0.20 0.41 0.10
5
2.6
0.50
0.00
Significance level UU UU
Group 1sautoclave workers; Group 2 s maintenance workers in the polymerisation plant; Group 3s workers in polyvinyl chloride production but not with autoclaves; Group 4 s maintenance workers in the same area as Group 3. UU ps - 0.01 by comparison with controls.
generalised linear model was used for equation fitting. For the petroleum emissions study, a t-test and analysis of variance ŽANOVA. were applied to determine whether there was a significant difference between one or more confounding factors and different variables Ža significance level lower than 0.05 is reported as U , and a level lower than 0.005 is reported as UU ..
3. Results Tables 1 and 2 show the results for the vinyl chloride study.
Fig. 1. Influence of exposure to petroleum emissions on biomarkers in the study population. Abbreviations: TAbF, total aberration frequency; AbC, percent of aberrant cells; SCE, sister chromatid exchange; HFC, percent of cells containing high frequency; RAS, ras p21 protein level. UU ps - 0.005.
Table 1 shows the % B cells Žgaps and breaks., % Cu cells Žrings, fragments and dicentrics. and % Cs cells Žtranslocations. in autoclave operators ŽGroup 1., maintenance workers who worked in the polymerisation plant ŽGroup 2., those who worked in polyvinyl chloride production but not with autoclaves ŽGroup 3., and Group 4 were maintenance workers in the same area as Group 3. Groups 1 and 2 had the highest exposure. The exposed groups, with increasing exposure showed a dose–response relationship for B and Cu cells, but for Cs cells, where only translocations but not inversions were scored Ždue to the staining techniques used., there was an increase above control values. Values were statistically significantly different from controls Ž p - 0.01. for maintenance workers and autoclave operators. When the values for Cu and Cs cells Ž% C cells. were tabulated ŽTable 2. with duration of employ-
Table 2 Correlation of individuals % C cells with three other parameters Duration of employment Control 1–10 years )10 years
0.58 1.59 1.92
Smoking history No smoking history Present non-smoker Present smoker
1.21 1.06 1.80
Excursion exposure No excursions Recent exposure excursions
1.46 2.29
Fig. 2. Influence of years of exposure to petroleum emissions on biomarkers in the study population. U ps - 0.05; UU ps - 0.005.
200
D. Andersonr Mutation Research 428 (1999) 197–202
Fig. 3. Influence of smoking on biomarkers in smokers and non-smokers in the study population. PRI, proliferative rate index. UU U ps - 0.05; ps - 0.005.
ment, smoking history or excursion exposure, Žan individual being able to smell vinyl chloride. there were clear relationships with these parameters. This tabulation illustrates that these three factors are associated with an increase in the percentage of C cell abnormalities. The duration of employment variable was significant at the 5% level, the present smoking variable was significant at the 10% level and the history of exposure to excursion levels was not significant. There was no positive correlation of chromosomal damage with various other parameters such as liver function tests Žbilirubin, platelets, g-glutamyl transpeptidase ŽGGT., alanine transaminase ŽALT., alkaline phosphatase ŽALP. and aspartame transaminase ŽAST. Ždata not shown.. The off-site controls did not undergo liver function tests.
Fig. 4. Influence of the extent of smoking on biomarkers in the study population. UU ps - 0.005.
Fig. 5. Influence of sex on biomarkers in the study population. U ps - 0.05.
Figs. 1–6 show the results for the petroleum emissions study. Fig. 1 shows the effect of exposure to petroleum emissions of workers by comparison with unexposed controls. There were statistically significant increases in chromosome aberration frequencies and aberrant cells in the exposed group. The influence of exposure to petroleum emissions on various biomarkers can be seen in Fig. 2, the influence of smoking in Figs. 3 and 4, gender in Fig. 5 and season in Fig. 6. There was also an effect of exposure on various chromosome biomarkers in the non-smoking population and in the unexposed sub-groups Ždata not shown..
Fig. 6. Seasonal influence on biomarkers in the study population. UU ps - 0.005.
D. Andersonr Mutation Research 428 (1999) 197–202
4. Discussion The relationship between the levels of chromosomal abnormalities and other biomarkers and the various factors which influence them is complex. For the vinyl chloriderpolyvinyl chloride workers there were factors which have been shown to be associated with increased chromosomal abnormalities: the job category, the length of employment, experience of exposure to short-term exposure levels and smoking history. It was not possible to estimate which of these four variables was the most important since the statistical analysis was complicated by the interrelationships between the variables and this is exemplified in Table 1, where three of the variables were shown to influence the level of C cell abnormalities. There was no positive correlation of chromosomal abnormalities with various other parameters Žbilirubin, platelets, GGT, ALP, ALT and AST. These liver function tests were investigated because vinyl chloride is known to be associated with angiosarcoma of the liver and it was thought there might be some association between chromosome damage and this cancer. With regard to the workers exposed to petroleum emissions, our results have revealed a significant increase in cytogenetic aberrations and ras p21 oncoproteins in the whole population studied Žincluding smokers and non-smokers. and also in non-smoking individuals. However, in contrast to exposure to vinyl chloride, for workers exposed to petroleum emissions, the increase was not correlated with the duration of exposure. Other studies have also reported no correlation between the duration of exposure to petroleum emissions w12,13x. Smoking was found to affect various biomarkers both in the whole population under study and in the unexposed group, as has previously been shown w14x. A seasonal influence was also observed where responses were higher in winter. Environmental pollution is higher in the winter season in Poland due to the intensive combustion of coal for residential heating during the winter months. A seasonal influence has been previously observed in a Polish population w14x and in a British population w15x. Not only are there confounding effects as shown in the above two studies, but there can also be further confounders if information about humans is
201
obtained indirectly. This can be illustrated by reproductive studies with vinyl chloride. Personal interviews andror questionnaires are a primary source of data for monitoring programmes. In gathering information concerning reproductive events, studies based on husbands’ indirect reports yielded considerably lower figures for pregnancy loss w16x than those based on interviews with wives w17x. Individuals clearly have a much better recall for events in their own lives, and the circumstances of pregnancy are far more significant for a woman than a man. Therefore, gathering information directly from the wives of employees would be a valuable technique in industrial reproductive monitoring programmes. For chromosome studies, however, the individual is generally interviewed directly or completes hisrher own questionnaire. The above studies have illustrated how complex the relationship is between the biomarkers investigated and the various factors which can influence them. This is because some of the variables are not completely independent of each other. References w1x A. Pope, The Essay on Man, Epistle 2, line 2 Ž1733r1734.. w2x D. Anderson, E. Zeiger, Human monitoring, Environ. Mol. Mutagen. 30 Ž1997. 95–96. w3x I.M. Tough, K.E. Buckton, A.G. Baikie, W.M. Court-Brown, X-ray induced chromosome damage in man, Lancet ii Ž1960. 849–851. w4x M.A. Bender, P.A. Gooch, Persistent chromosome aberrations in irradiated human subjects, Radiat. Res. 16 Ž1962. 44–53. w5x I.M. Tough, W.M. Court-Brown, Chromosome aberrations and exposure to ambient benzene, Lancet i Ž1965. 684. w6x Nordic Study Group on the Health Risk of Chromosome Damage, L. Hagmar, A. Brogger, I.L. Hansteen, S. Heim, B. Hogstedt, L. Knudsen, B. Lambert, K. Linnainmaa, F. Mitelman, I. Nordenson, C. Reuterwall, S. Salomaa, S. Skerfvuing, M. Sorsa, Cancer risk in humans predicted by increased levels of chromosomal aberrations in lymphocytes, Cancer Res. 54 Ž1994. 2919–2922. w7x S. Bonassi, A. Abbondandolo, L. Camurri, L. DalPra, F. De Ferrari, F. Degrassi, A. Forni, L. Lamberti, C. Lando, P. Padovani, I. Sbrana, D. Vecchio, R. Puntoni, Are chromosome aberrations in circulating lymphocytes predictive of future cancer onset in humans? Preliminary results of an Italian cohort study, Cancer Genet. Cytogenet. 79 Ž1995. 133–135. w8x A. Forni, L. Moreo, Cytogenetic studies in a case of benzene leukaemia, Eur. J. Cancer 3 Ž1967. 251–255.
202
D. Andersonr Mutation Research 428 (1999) 197–202
w9x I.F.H. Purchase, C.R. Richardson, D. Anderson, Chromosome effects in peripheral lymphocytes, Proc. R. Soc. Med. 69 Ž1976. 290–291. w10x I.F.H. Purchase, C.R. Richardson, D. Anderson, G.M. Paddle, W.G.F. Adams, Chromosomal analysis in vinyl chloride exposed workers, Mutat. Res. 57 Ž1978. 325–334. w11x D. Anderson, J.A. Hughes, A. Cebulska-Wasilewska, A. Wierzewska, E. Kasper, Biological monitoring of workers exposed to emissions from petroleum plants, Environ. Health Perspect. 104 Ž1996. 609–613. w12x A. Tompa, J. Major, M.G. Jakab, Monitoring of benzene exposed workers for genotoxic effects of benzene, Mutat. Res. 304 Ž1994. 159–165. w13x J. Major, M.G. Jakab, G. Kiss, A. Tompa, Chromosome aberration, sister-chromatid exchange, proliferative rate index and serum thiocyanate concentration in smokers exposed to low-dose benzene, Environ. Mol. Mutagen. 23 Ž1994. 137– 142.
w14x F. Perera, K. Hemminki, E. Gryzbowska, G. Motykiewics, J. Michalska, R.M. Santella, T.L. Young, C. Dickey, P. Brandt-Rauf, I. DeVivo, W. Blaner, W.-Y. Twai, M. Chorazy, Molecular and genetic damage in humans from environmental pollution in Poland, Nature 360 Ž1992. 256–258. w15x D. Anderson, A.J. Francis, P. Godbert, P.C. Jenkinson, K.R. Butterworth, Chromosome aberrations ŽCA., sister chromatid exchanges ŽSCE. and mitogen induced blastogenesis in cultured peripheral lymphocytes from 48 individuals sampled 8 times over 2 years, Mutat. Res. 250 Ž1991. 467–476. w16x P.F. Infante, J.K. Wagoner, A.J. McMichael, R.J. Waxweiler, H. Falk, Genetic risks of vinyl chloride, Lancet Ž1976., 734–735. w17x P.A. Buffler, J.M. Aase, Genetic risks and environmental surveillance, J. Occup. Med. 24 Ž1982. 305–314. w18x K.E. Buckton, M.C. Pike, Time in culture: an important variable in studying in vivo radiation, Int. J. Radiat. Biol. 8 Ž1964. 439–452.
Factors contributing to biomarker responses in exposed workers Diana Anderson
)
BIBRA International, Woodmansterne Road, Carshalton, Surrey SM5 4DS, UK Received 29 November 1998; accepted 11 January 1999
Abstract ‘‘ . . . wThxe proper study of mankind is man’’ ŽPope, circa 1733r1734.. Human monitoring fits this notion and monitoring after exposure to genotoxic agents is now an established discipline. It is possible in many situations to identify humans exposed to potentially toxic materials in the workplace and the environment. Responses are often measured in peripheral lymphocytes because these cells can be acquired by a generally socially and ethically acceptable, minimally invasive route. In the early 1960s, chromosome damage in these cells was one of the first endpoints to be used as a biomarker and benzene was one of the first chemicals investigated. Although a causal relationship between chromosome damage and cancer has not been proven, it has been suggested to have some prognostic significance for future cancer onset. With other genetic biomarkers this is as yet not the case, but there are now many biomarkers for different areas of toxicology. Other well-established genetic biomarkers include the detection of hprt mutations, micronuclei and sister chromatid exchanges. However, for interpretation of responses, the issue of confounding factors must be addressed. As in most human studies, there tends to be a high degree of interindividual variability in response to chemical insults. Some non-exposed control individuals exhibit as high a level of damage as some exposed individuals and some of these have levels of damage as low as many of the controls. Thus, it is only the mean values of the groups that can substantiate an exposure related-problem; the data on an individual basis are still of limited use. While human lymphocytes remain the most popular cell type for monitoring purposes, sperm, buccal, nasal, epithelial and placental cells are also used. Confounding factors affect responses in all cell types. There are endogenous confounding factors such as age, sex, genetic make-up and exogenous confounding factors including lifestyle habits such as smoking, drinking, etc. There are biomarkers of exposure, effect and susceptibility and the last may be influenced by the genotype and polymorphism genes existing in a population. From our own studies, confounding effects will be considered in relation to workers exposed to vinyl chloride and petroleum emissions. The relationship between the biomarkers and various factors which influence them is complex. Sometimes the variables are not completely independent of one another. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Confounding factor; Age; Sex; Smoking; Drinking; Lymphocyte; Vinyl chloride; Petroleum emission
1. Introduction ‘‘The proper study of mankind is man’’ w1x. Human monitoring fits this notion and monitoring after ) Tel.: q44-181-6521000; Fax: q44-181-6617029; E-mail: [email protected]
exposure to genotoxic agents is now an established discipline w2x. Human somatic and germinal mutation frequencies may be increased by exposure to a variety of agents. Genic or chromosomal somatic mutations may contribute to acquired disorders such as cancer and germinal mutations and are likely to contribute
0027-5107r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 4 2 Ž 9 9 . 0 0 0 4 7 - 2
198
D. Andersonr Mutation Research 428 (1999) 197–202
to inherited defects in the offspring of individuals exposed to mutagenic agents. There is thus a continuing need to develop and supply methods to study exposed populations at risk. It is possible in many situations to identify humans exposed to potentially toxic materials in the workplace and the environment. Responses are often measured in peripheral lymphocytes because these cells can be acquired by a generally socially and ethically acceptable, minimally invasive route. Radiation was one of the first agents to be examined in this way in the early 1960s and chromosome damage was one of the first endpoints to be used w3,4x. Chemical damage was similarly measured in the mid-1960s w5x. Although a causal relationship between chromosomal aberrations and cancer has not been proven, a prospective study by the Nordic Group that examined 3182 individuals w6x showed a highly statistically significant linear trend for a positive association Ž p s 0.0009. between chromosome damage in peripheral lymphocytes and subsequent cancer risk. The data were not extensive enough to determine if the same associations existed for sister chromatid exchanges and micronuclei. Bonassi et al. w7x also suggested that chromosomal aberrations in circulating lymphocytes were predictive of future cancer onset in humans. Furthermore, many studies have shown a higher level of cytogenetic damage in tumour cells than in normal cells, and an early association of this nature was made by Forni and Moreo w8x. Methods to detect gene mutations have also been long established Že.g., for the HPRT locus. and increased mutation frequencies have been shown to result from genotoxic exposures. Some of the newer monitoring approaches include determination of cytogenetic damage using fluorescence in situ hybridisation ŽFISH. techniques, metabolic genotypes that might show altered sensitivity, such as GSTMI and NAT2 and DNA strand breakage in the single cell gel DNA electrophoresis ŽComet assay.. DNA microsatellite instabilities, oncoproteins Že.g., ras p21. and tumour suppressor genes Žp53. are also involved in the cancer process, and measurement of these has now been added to the monitoring armoury. While human lymphocytes remain the most popular cell type for monitoring purposes, sperm, buccal, nasal epithelial and placental cells are also being used.
These cells can be acquired with minimal invasion of the individual. Exposure to genotoxic agents can result from natural and environmental factors, non-specific contamination, the occupational environment, or industrial accidents. For interpretation of responses, the issue of confounding factors must be addressed. As in most human studies, there tends to be a high degree of interindividual variability in response to toxic insults. Some non-exposed individuals exhibit as high a level of damage as some exposed individuals and some of those exposed have levels of damage as low as many of the controls. Thus, it still pertains that only the mean values of the groups can substantiate an exposure-related problem; the data on an individual basis are still of limited use. Confounding factors affect responses in all cell types. There are endogenous confounding factors such as age, sex, genetic make-up and exogenous confounding factors include lifestyle habits such as smoking, drinking, etc. There are biomarkers of exposure, effect and susceptibility and the last may be influenced by the genotype and polymorphisms in genes existing in a population. From our own studies, confounding effects will be considered in relation to workers exposed to vinyl chloride and petroleum emissions.
2. Materials and methods 2.1. Sampling Interviews were performed using a questionnaire. Blood samples were collected from exposed and unexposed groups. To avoid a seasonal influence, sampling of exposed and unexposed groups was carried out during the same time period. The methods used for the preparation of samples and the details of the studies described in the present communication have been published elsewhere w9– 11x. 2.2. Statistics For the vinyl chloride study, the results were assumed to conform to a Poisson distribution and the
D. Andersonr Mutation Research 428 (1999) 197–202
199
Table 1 Chromosomal abnormalities in workers employed in polyvinylchloride manufacture ŽB and C cells classified according to Buckton and Pike w18x.
Group 1 Group 2 Group 3 Group 4 Controls on-site Controls off-site
No. of workers
%B cells
% Cu cells
% Cs cells
18 21 5 12 19
6.4 6.9 6.6 5.1 3.6
1.72 1.48 1.20 1.10 0.53
0.44 0.30 0.20 0.41 0.10
5
2.6
0.50
0.00
Significance level UU UU
Group 1sautoclave workers; Group 2 s maintenance workers in the polymerisation plant; Group 3s workers in polyvinyl chloride production but not with autoclaves; Group 4 s maintenance workers in the same area as Group 3. UU ps - 0.01 by comparison with controls.
generalised linear model was used for equation fitting. For the petroleum emissions study, a t-test and analysis of variance ŽANOVA. were applied to determine whether there was a significant difference between one or more confounding factors and different variables Ža significance level lower than 0.05 is reported as U , and a level lower than 0.005 is reported as UU ..
3. Results Tables 1 and 2 show the results for the vinyl chloride study.
Fig. 1. Influence of exposure to petroleum emissions on biomarkers in the study population. Abbreviations: TAbF, total aberration frequency; AbC, percent of aberrant cells; SCE, sister chromatid exchange; HFC, percent of cells containing high frequency; RAS, ras p21 protein level. UU ps - 0.005.
Table 1 shows the % B cells Žgaps and breaks., % Cu cells Žrings, fragments and dicentrics. and % Cs cells Žtranslocations. in autoclave operators ŽGroup 1., maintenance workers who worked in the polymerisation plant ŽGroup 2., those who worked in polyvinyl chloride production but not with autoclaves ŽGroup 3., and Group 4 were maintenance workers in the same area as Group 3. Groups 1 and 2 had the highest exposure. The exposed groups, with increasing exposure showed a dose–response relationship for B and Cu cells, but for Cs cells, where only translocations but not inversions were scored Ždue to the staining techniques used., there was an increase above control values. Values were statistically significantly different from controls Ž p - 0.01. for maintenance workers and autoclave operators. When the values for Cu and Cs cells Ž% C cells. were tabulated ŽTable 2. with duration of employ-
Table 2 Correlation of individuals % C cells with three other parameters Duration of employment Control 1–10 years )10 years
0.58 1.59 1.92
Smoking history No smoking history Present non-smoker Present smoker
1.21 1.06 1.80
Excursion exposure No excursions Recent exposure excursions
1.46 2.29
Fig. 2. Influence of years of exposure to petroleum emissions on biomarkers in the study population. U ps - 0.05; UU ps - 0.005.
200
D. Andersonr Mutation Research 428 (1999) 197–202
Fig. 3. Influence of smoking on biomarkers in smokers and non-smokers in the study population. PRI, proliferative rate index. UU U ps - 0.05; ps - 0.005.
ment, smoking history or excursion exposure, Žan individual being able to smell vinyl chloride. there were clear relationships with these parameters. This tabulation illustrates that these three factors are associated with an increase in the percentage of C cell abnormalities. The duration of employment variable was significant at the 5% level, the present smoking variable was significant at the 10% level and the history of exposure to excursion levels was not significant. There was no positive correlation of chromosomal damage with various other parameters such as liver function tests Žbilirubin, platelets, g-glutamyl transpeptidase ŽGGT., alanine transaminase ŽALT., alkaline phosphatase ŽALP. and aspartame transaminase ŽAST. Ždata not shown.. The off-site controls did not undergo liver function tests.
Fig. 4. Influence of the extent of smoking on biomarkers in the study population. UU ps - 0.005.
Fig. 5. Influence of sex on biomarkers in the study population. U ps - 0.05.
Figs. 1–6 show the results for the petroleum emissions study. Fig. 1 shows the effect of exposure to petroleum emissions of workers by comparison with unexposed controls. There were statistically significant increases in chromosome aberration frequencies and aberrant cells in the exposed group. The influence of exposure to petroleum emissions on various biomarkers can be seen in Fig. 2, the influence of smoking in Figs. 3 and 4, gender in Fig. 5 and season in Fig. 6. There was also an effect of exposure on various chromosome biomarkers in the non-smoking population and in the unexposed sub-groups Ždata not shown..
Fig. 6. Seasonal influence on biomarkers in the study population. UU ps - 0.005.
D. Andersonr Mutation Research 428 (1999) 197–202
4. Discussion The relationship between the levels of chromosomal abnormalities and other biomarkers and the various factors which influence them is complex. For the vinyl chloriderpolyvinyl chloride workers there were factors which have been shown to be associated with increased chromosomal abnormalities: the job category, the length of employment, experience of exposure to short-term exposure levels and smoking history. It was not possible to estimate which of these four variables was the most important since the statistical analysis was complicated by the interrelationships between the variables and this is exemplified in Table 1, where three of the variables were shown to influence the level of C cell abnormalities. There was no positive correlation of chromosomal abnormalities with various other parameters Žbilirubin, platelets, GGT, ALP, ALT and AST. These liver function tests were investigated because vinyl chloride is known to be associated with angiosarcoma of the liver and it was thought there might be some association between chromosome damage and this cancer. With regard to the workers exposed to petroleum emissions, our results have revealed a significant increase in cytogenetic aberrations and ras p21 oncoproteins in the whole population studied Žincluding smokers and non-smokers. and also in non-smoking individuals. However, in contrast to exposure to vinyl chloride, for workers exposed to petroleum emissions, the increase was not correlated with the duration of exposure. Other studies have also reported no correlation between the duration of exposure to petroleum emissions w12,13x. Smoking was found to affect various biomarkers both in the whole population under study and in the unexposed group, as has previously been shown w14x. A seasonal influence was also observed where responses were higher in winter. Environmental pollution is higher in the winter season in Poland due to the intensive combustion of coal for residential heating during the winter months. A seasonal influence has been previously observed in a Polish population w14x and in a British population w15x. Not only are there confounding effects as shown in the above two studies, but there can also be further confounders if information about humans is
201
obtained indirectly. This can be illustrated by reproductive studies with vinyl chloride. Personal interviews andror questionnaires are a primary source of data for monitoring programmes. In gathering information concerning reproductive events, studies based on husbands’ indirect reports yielded considerably lower figures for pregnancy loss w16x than those based on interviews with wives w17x. Individuals clearly have a much better recall for events in their own lives, and the circumstances of pregnancy are far more significant for a woman than a man. Therefore, gathering information directly from the wives of employees would be a valuable technique in industrial reproductive monitoring programmes. For chromosome studies, however, the individual is generally interviewed directly or completes hisrher own questionnaire. The above studies have illustrated how complex the relationship is between the biomarkers investigated and the various factors which can influence them. This is because some of the variables are not completely independent of each other. References w1x A. Pope, The Essay on Man, Epistle 2, line 2 Ž1733r1734.. w2x D. Anderson, E. Zeiger, Human monitoring, Environ. Mol. Mutagen. 30 Ž1997. 95–96. w3x I.M. Tough, K.E. Buckton, A.G. Baikie, W.M. Court-Brown, X-ray induced chromosome damage in man, Lancet ii Ž1960. 849–851. w4x M.A. Bender, P.A. Gooch, Persistent chromosome aberrations in irradiated human subjects, Radiat. Res. 16 Ž1962. 44–53. w5x I.M. Tough, W.M. Court-Brown, Chromosome aberrations and exposure to ambient benzene, Lancet i Ž1965. 684. w6x Nordic Study Group on the Health Risk of Chromosome Damage, L. Hagmar, A. Brogger, I.L. Hansteen, S. Heim, B. Hogstedt, L. Knudsen, B. Lambert, K. Linnainmaa, F. Mitelman, I. Nordenson, C. Reuterwall, S. Salomaa, S. Skerfvuing, M. Sorsa, Cancer risk in humans predicted by increased levels of chromosomal aberrations in lymphocytes, Cancer Res. 54 Ž1994. 2919–2922. w7x S. Bonassi, A. Abbondandolo, L. Camurri, L. DalPra, F. De Ferrari, F. Degrassi, A. Forni, L. Lamberti, C. Lando, P. Padovani, I. Sbrana, D. Vecchio, R. Puntoni, Are chromosome aberrations in circulating lymphocytes predictive of future cancer onset in humans? Preliminary results of an Italian cohort study, Cancer Genet. Cytogenet. 79 Ž1995. 133–135. w8x A. Forni, L. Moreo, Cytogenetic studies in a case of benzene leukaemia, Eur. J. Cancer 3 Ž1967. 251–255.
202
D. Andersonr Mutation Research 428 (1999) 197–202
w9x I.F.H. Purchase, C.R. Richardson, D. Anderson, Chromosome effects in peripheral lymphocytes, Proc. R. Soc. Med. 69 Ž1976. 290–291. w10x I.F.H. Purchase, C.R. Richardson, D. Anderson, G.M. Paddle, W.G.F. Adams, Chromosomal analysis in vinyl chloride exposed workers, Mutat. Res. 57 Ž1978. 325–334. w11x D. Anderson, J.A. Hughes, A. Cebulska-Wasilewska, A. Wierzewska, E. Kasper, Biological monitoring of workers exposed to emissions from petroleum plants, Environ. Health Perspect. 104 Ž1996. 609–613. w12x A. Tompa, J. Major, M.G. Jakab, Monitoring of benzene exposed workers for genotoxic effects of benzene, Mutat. Res. 304 Ž1994. 159–165. w13x J. Major, M.G. Jakab, G. Kiss, A. Tompa, Chromosome aberration, sister-chromatid exchange, proliferative rate index and serum thiocyanate concentration in smokers exposed to low-dose benzene, Environ. Mol. Mutagen. 23 Ž1994. 137– 142.
w14x F. Perera, K. Hemminki, E. Gryzbowska, G. Motykiewics, J. Michalska, R.M. Santella, T.L. Young, C. Dickey, P. Brandt-Rauf, I. DeVivo, W. Blaner, W.-Y. Twai, M. Chorazy, Molecular and genetic damage in humans from environmental pollution in Poland, Nature 360 Ž1992. 256–258. w15x D. Anderson, A.J. Francis, P. Godbert, P.C. Jenkinson, K.R. Butterworth, Chromosome aberrations ŽCA., sister chromatid exchanges ŽSCE. and mitogen induced blastogenesis in cultured peripheral lymphocytes from 48 individuals sampled 8 times over 2 years, Mutat. Res. 250 Ž1991. 467–476. w16x P.F. Infante, J.K. Wagoner, A.J. McMichael, R.J. Waxweiler, H. Falk, Genetic risks of vinyl chloride, Lancet Ž1976., 734–735. w17x P.A. Buffler, J.M. Aase, Genetic risks and environmental surveillance, J. Occup. Med. 24 Ž1982. 305–314. w18x K.E. Buckton, M.C. Pike, Time in culture: an important variable in studying in vivo radiation, Int. J. Radiat. Biol. 8 Ž1964. 439–452.