- Email: [email protected]
Polychlorinated biphenyls contamination in women with breast cancer
Clinica Chimica Acta 347 (2004) 177 – 181 www.elsevier.com/locate/clinchim
Polychlorinated biphenyls contamination in women with breast cancer Corinne J. Charlier a,*, Adelin I. Albert b, Liying Zhang b, Nathalie G. Dubois a, Guy J. Plomteux a a
Clinical Toxicology Laboratory, Liege University Hospital, Tour II+5, CHU Sart-Tilman, B 4000 Lie`ge, Belgium b Biostatistics Department, University of Liege, Sart Tilman, B 35, 4000 Liege, Belgium Received 9 March 2004; received in revised form 15 April 2004; accepted 19 April 2004
Abstract Background: Polychlorinated biphenyls (PCBs) are widespread highly resistant pollutants in the environment with potential adverse health effects on humans. The aim of the study was to compare PCBs contamination in women suffering from breast cancer with presumably healthy women. Methods: A gas-chromatography/mass-spectrometry method was used to identify and quantify seven PCBs congeners (IUPAC 28, 52, 101, 118, 138, 153,180) in blood from 60 cases of breast cancer and 60 agematched healthy controls. Results: Cases and controls had similar risk profiles, except for menopausal status (respectively 82% vs. 65%, p = 0.014). PCBs were detectable in 69.1% of the samples. Total blood level of PCBs was significantly different ( p = 0.012) in cases (7.08 F 7.51 ppb) and controls (5.10 F 5.15 ppb). The relationship between PCBs concentrations in serum and risk factor was mainly due to serum levels of PCB153, which were significantly higher in breast cancer women than in disease-free subjects (1.63 F 1.26 ppb vs. 0.63 F 0.78 ppb, p < 0.0001), even after accounting for other potential risk factors. Conclusions: These results suggest that environmental exposure to PCBs may contribute to multifactorial pathogenesis of breast cancer. D 2004 Elsevier B.V. All rights reserved. Keywords: Breast cancer; Endocrine disrupters; PCBs
1. Introduction Polychlorinated biphenyls (PCBs) are complex chemical mixtures that comprise theoretically 209 congeners, some of which are known to cause a wide range of adverse effects on animals and humans [1,2]. PCBs are industrial chemicals that have been used for
* Corresponding author. Tel.: +32-4-3668818; fax: +32-43668889. E-mail address: [email protected] (C.J. Charlier). 0009-8981/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.cccn.2004.04.025
diverse commercial applications such as hydraulic fluids, printing inks, or dielectric fluids for capacitors. The production and use of these compounds were banned in the late 1970s. Together with organochlorine pesticides, PCBs constitute what is called Persistent Organic Pollutants (POPs), due to their great chemical stability, their lipid solubility, and their ubiquitous prevalence in environment [3,4]. Adults are mainly exposed through the consumption of dairy products, meat and fish. The presence of POPs in human serum and adipose tissue has been reported in many studies over the last three decades. Several
178
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
epidemiological studies have suggested that exposure to POPs may increase the risk of endocrine disruption by interference with estrogen metabolism [5 – 7], elevation of the amount of bioavailable estradiol [8], or synergistic effects on estrogen responsive genes [9]. POPs have also been hypothesized to be associated with a variety of effects including hypofertility in man [10], carcinogenesis in man (prostate, testis) [11], immunological perturbations and neurological effects [12], as well as breast cancer in women [13 – 19]. Results concerning breast carcinogenesis are controversial. Despite marked regional differences in prevalent cases (high in North America, Western Europe and Oceania, low in Africa and Asia), the cause of most breast cancers is not yet known. Genetics explains about 10% of the cases. Risk factors for breast cancer include the reproductive life of women (early menarche, nulliparity or late age at first birth, late menopause) as well as hormonal factors, whether endogenous (high levels of estrogens, free or bound to sex hormone binding globulin or exogenous [long term use of hormonal replacement therapy, (HRT)]. The exact role of exposure to other products with implication on the hormonal activity, as may be the case for PCBs, is not yet known. In the present study, we compared the blood levels of seven frequently detected PCBs congeners (IUPAC no. 28, 52, 101, 118, 138, 153 and 180) in women with breast cancer and in non-diseased female controls of the same age. Pollutants were quantified simultaneously using a gas chromatographic analyzer coupled to an ion-trap mass spectrometer detector. The PCBs serum levels observed in cases and controls were compared, while taking into account other risk factors for breast cancer.
2. Material and methods 2.1. Study populations The present study involved 60 women, aged 54.8 F 6.59 years, diagnosed with breast cancer and undergoing a surgical intervention (cases), and 60 control women of the same mean age (54.8 F 6.55 years) selected at random in a population of presumably healthy women consulting for routine systematic cervico-vaginal cytological screening. For each sub-
ject, information regarding age at menarche, pregnancy, breast-feeding, menopausal status and family history of breast cancer (mother, sister or grandmother) were recorded by interview. Postmenopausal women were asked for use of HRT. All patients gave their informed consent for participating in the study. For controls, blood specimens were taken at the time of examination, whereas for cases samples were collected at the time of diagnosis prior to surgery. Blood samples (10 ml) were drawn in the early morning after overnight fasting and were immediately centrifuged with serum specimens kept frozen at 18 jC until assay (within 1 week). 2.2. PCB analysis The quantification of the seven PCBs in serum was done using a gas chromatographic analyser coupled to an ion-trap mass spectrometer detector (Saturn 2000, Varian) by a method described elsewhere [20]. Briefly, calibration samples or serum samples (1 ml) were extracted with 1 ml acetonitrile, 1 ml saturated K2CO3 and 5 ml petroleum ether/diethyl ether (98:2, v/v) after addition of 10 Al of internal standard (aldrin, 10 mg/l). The samples were vortex mixed, then centrifuged (2500 rpm, 5 min) and the supernatant was transferred to a glass collection vial. A repeated extraction with another 5 ml of petroleum ether/ diethyl ether was performed and the supernatant added to the first one. The two fractions were simultaneously transferred to the pre-conditioned extraction column. The solution drawn through the cartridge was 3collected and evaporated under mild nitrogen at 35 jC. The residue was reconstituted in 100 Al hexane and 1 Al was injected in split/splitless (ratio 1:35) mode onto the gas chromatograph. The samples were eluted through a HP5 Trace (30 m, 0.25 mm id, 0.25 Am film) column (Agilent). For detection, the mass spectrometry was operated in the electronic impact mode (70 eV). Selected ion storage was used for identification and quantification. All reagents were of the highest purity available. All PCBs and internal standard (aldrin) were purchased from Dr. Ehrenstorfer (Ausburg, Germany). Samples were analysed in a blind procedure together with controls consisting of samples spiked with 0.5 or 2 ppb of each PCB. Interlaboratory comparison program (AMAP Ring
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
test, Institut National Sante´ Publique, Que´bec) was performed as external quality control.
Table 1 Characteristics of risk factors in women with breast cancer (cases) and in disease-free women (controls) of similar age
2.3. Statistical analysis Serum levels of PCBs were expressed as mean F SD. Correction for lipid content was tested, but since the results were not affected, only crude data are presented. Quantitative continuous results were not categorized but a log-transform was applied to PCBs data to normalize their distributions. All subsequent statistical analyses were carried out on the transformed data. Risk factors and PCBs serum levels in cases and controls were compared by means of ordinary logistic regression, not only for each congener separately but also for all PCBs simultaneously. In every comparison, age was included in the analysis. Results were expressed as odds ratios (OR) and associated 95% confidence intervals (CI). In the multivariate analysis, adjustment was made for potential confounding risk factors: menopausal status, number of full-term pregnancies, lactation, use of HRT and family history of breast cancer. Results were confirmed after backward elimination of not significant effects. All results were considered to be significant at the 5% critical level ( p < 0.05). Calculations were done using the SAS (SAS Institute, version 8.2 for Windows) and S-Plus (version 6.2) statistical packages.
3. Results and discussion 3.1. Risk factors The characteristics of cases and controls are displayed in Table 1. The two groups had similar risk profiles but the prevalence of menopause was significantly higher in women with breast cancer, yielding to an odds ratio of 3.8 (95% CI: 1.3 – 11).
179
Age of menarche, mean years (SD) Menopause, n (%) HRTb, n (%) Parity, n (%) Breast feedingc, n (%) Family history of breast cancer, n (%)
Cases (n = 60)
Controls (n = 60)
pa
11.4 (0.95)
11.5 (1.08)
0.59
49 30 34 24 9
(81.7) (61.2) (56.7) (70.6) (15.0)
39 26 33 25 5
(65.0) (66.7) (55.0) (75.6) (8.30)
0.014 0.45 0.85 0.85 0.25
a
p-value adjusted for age. Restricted to menopaused women. c Restricted to parous women. b
were significantly higher in cases than in controls. The same observation was made for total PCB content ( p = 0.012). 3.3. Association with breast cancer All PCB concentrations and confounding risk factors for breast cancer were combined into a multiple logistic regression analysis (see Table 3). It is seen that menopause ( p = 0.038) and PCB153 ( p < 0.0001) remained statistically discriminate between cases and controls while the effect of PCB138 disappeared. Thus, high concentrations of PCB153 are significantly associated with an increased risk of breast cancer despite the presence of other factors, menopause in particular. After backward elimination of not significant effects, menopause and serum PCB 153 concentration remained significant (respectively, p = 0,036 and p < 0,0001), reinforcing the previous conclusion. Similar results were obtained when replacing individual PCBs by total PCB content. The only two significant variables were menopausal status ( p = 0.015) and total serum PCB concentration ( p = 0.035). 3.4. Discussion
3.2. PCB concentrations PCBs 28 and 118 could not be detected in any of the subjects of the study. The distributions of the five other congeners in cases and controls are presented in Table 2. For PCBs 52, 101 and 180, serum concentrations did not differ in the two groups. By contrast, the serum concentrations of PCB138 and PCB153
In the present case-control study, PCB153, as already suggested by a previous limited study [21] and, to a lesser extent, PCB138 were found in significantly higher concentration in blood samples of women diagnosed with breast cancer when compared to a control group of disease free subjects of the same age. In our study, serum PCB153 concentration
180
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
Table 2 Means (SD) of PCBsa serum levels (ppb) in women with breast cancer (cases) and in disease-free women (controls) of similar age Congener
Cases (n = 60)
Controls (n = 60)
pb
PCB52 PCB101 PCB138 PCB153 PCB180 Total PCB content
1.64 0.77 1.25 1.63 1.79 7.08
1.67 0.62 0.94 0.63 1.24 5.10
0.87 0.23 0.0068 < 0.0001 0.17 0.012
(3.45) (1.21) (1.20) (1.26) (3.46) (7.51)
(3.06) (1.38) (1.84) (0.78) (1.45) (5.15)
a
PCBs 28 and 118 were undetected. p-values were obtained by logistic regression applied to log (PCB) after adjusting for age. b
remained associated with an increased risk of breast cancer, even after accounting for classical breast cancer risk factors. In a previously published paper, the PCB153 level correlated with the sum of other PCBs in blood and was proposed as an indicator substance in monitoring studies [22]. In another recent study [23], congeners 118, 138, 153 and 180 were highly intercorrelated, but not correlated with congeners 28 or 101. Our results are in agreement with preeminence of PCB153 and with correlation between PCB153 and total serum PCB concentration. PCB153 is an ortho-substituted PCB congener, able to present some estrogenic activity such as increasing uterine weight, causing precocious puberty in rats, and enhancing proliferation of MCF-7 breast tumour cells [24]. Moreover, PCBs that have two orthosubstitutions together with two parasubstitutions have been described as phenobarbital-like cytochromes inducers [1,25] and PCB153 could induce oxidative DNA damage implicated in breast carcinogenesis [26]. It is yet unclear what relationship there may be between environmental exposure to these types of compounds and either the initiation or progression of certain diseases. However, these compounds, entering into the human body via the food chain, are characterized by a long half-life and lipophilic properties, which facilitate their accumulation in adipose tissues. They have been detected in human tissues such as blood, milk or fat [27,28]. Blood is the most accessible biological fluid and close correlation between the concentrations in blood and the total body burden can be obtained [29]. Bench top instrumentations have allowed the development of easy methods to initiate epidemiological studies in standardized conditions and
the possible association of PCBs with breast cancer has been postulated [30]. Depending on the reports, absence of correlation [31] but also association with one or another PCB have been reported [28]. In a population of 65 women treated by excision biopsy, PCB28 was found to be the most important risk factor (OR = 9.6) [30]. Association between PCB levels and breast cancer risk was sometimes restricted to subgroups (e.g. obese) of women [32]. In our study, PCB153 appeared to be the ‘‘specific’’ congener if an association with breast cancer risk is evaluated. Moreover, after accounting for potential confounding factors, PCB153 remained a significant risk factor. However, given the complexity of PCB mixtures in environmental exposures, it is reasonable to assume that different spectra of toxicity outcomes may be observed and we agree with other authors [23] that congeners to be measured in any specific epidemiological study will depend on the purpose of the study. In conclusion, our results comfort the debate that there is not sufficient evidence to answer the question on human risk resulting from low-dose endocrine-related effects. Further interdisciplinary research, combining detection and quantitation of pollutants, epidemiological data collection, but also metabolic polymorphism investigations should be conducted to identify individual compounds or pollutants associations that might be contributing to human cancer.
Table 3 Multiple logistic regression analysis applied to risk factors and PCBs concentrationsa in 60 women with breast cancer (cases) and 60 disease-free women (controls) of the same age Risk factors
OR (95%CI)
p
Age (years) 0.94 (0.85 – 1.02) 0.13 Age at menarche (years) 0.89 (0.58 – 1.38) 0.61 Menopause (yes) 4.4 (1.1 – 18) 0.038 HRT(yes)b 0.67 (0.24 – 1.9) 0.45 Parity (yes) 1.5 (0.4 – 5.8) 0.53 Breast feeding (yes)c 0.73 (0.19 – 2.7) 0.64 Family history of breast cancer (yes) 0.84 (0.21 – 3.3) 0.81 PCB52 (ppb) 0.95 (0.74 – 1.2) 0.70 PCB101 (ppb) 1.0 (0.77 – 1.3) 0.95 PCB138 (ppb) 1.2 (0.88 – 1.5) 0.29 PCB153 (ppb) 1.8 (1.4 – 2.5) < 0.0001 PCB180 (ppb) 1.1 (0.76 – 1.5) 0.64 a
A log-transform was used for all PCBs. Restricted to menopaused women. c Restricted to parous women. b
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
References [1] Brouwer A, Longnecker MP, Birnbaum LS, Cogliano J, Kostyniak P, Moore J, et al. Characterization of potential endocrine-related heath effects at low dose levels of exposure to PCBs. Environ Health Perspect 1999;107(Suppl. 4):639 – 49. [2] Mc Kinney JD, Waller CL. PCBs as hormonally active structural analogues. Environ Health Perspect 1994;102:290 – 7. [3] Chevreuil M, Granier L. Les PCBs, des polluants difficiles a` e´liminer. La recherche 1992;23:484 – 6. [4] Safe SH. PCBs: environmental impact, biochemical and toxic responses and implications for risk assessment. Crit Rev Toxicol 1994;24:87 – 149. [5] Korach KS, Sarver P, Chae K, Mc Lachlan JA, Mc Kinney JD. Estrogen receptor binding activity of polychlorinated hydroxy-biphenyls: conformationally restricted structural probes. Mol Pharmacol 1988;33:120 – 6. [6] Robinson AK, Sirbasku DA, Stancel GM. DDT supports the growth of an estrogen-responsive tumour. Toxicol Lett 1985;27:109 – 13. [7] Connor K, Ramamoorthy K, Moore M, Mustain M, Chen I, Safe S, et al. Hydroxylated PCBs as estrogens and antiestrogens: structure – activity relationship. Toxicol Appl Pharmacol 1997;142:160 – 8. [8] Gellert JR. Uterotrophic activity of PCBs and induction of precocious reproductive ageing in neonatally treated female rats. Environ Res 1978;16:123 – 30. [9] Jansen HT, Cookes PS, Porcelli J, Liu T-C, Hansen LG. Estrogenic and antiestrogenic actions of PCBs in female rats: in vitro and in vivo studies. Reprod Toxicol 1993;7: 237 – 48. [10] Oliva A, Spira A, Multigner L. Contribution of environmental factors to the risk of male infertility. Hum Reprod 2001;16: 1768 – 76. [11] Giwercan AE, Skakkebaek NE. The human testis: an organ at risk? Int J Androl 1992;15:373 – 85. [12] Shain W, Bush B, Seegal R. Neurotoxicity of PCBs: structure – activity relationship of individual congeners. Toxicol Appl Pharmacol 1991;111:33 – 42. [13] Cocco P. On the rumors about the silent spring. Review of the scientific evidence linking the occupational and environmental pesticide exposure to endocrine disruption health effects. Cad Sau`de Pu`blica 2002;18:379 – 402. [14] Laden F, Hankinson SE, Wolff MS, Colditz GA, Willett WC, Speizer FE, et al. Plasma organochlorine levels and the risk of breast cancer: an extended follow-up in the nurse’s health study. Int J Cancer 2001;91:568 – 74. [15] Aronson KJ, Miller AB, Woolcott CG, Sterns EE, McGready DR, Lickley LA, et al. Breast adipose tissue concentrations of polychlorinated biphenyls and other organochlorines and breast cancer risk. Cancer Epidemiol Biomarkers Prev 2000;9:55 – 63. [16] Davis DL, Bradlow HL, Wolff M, Woodruff T, Hoel DG, Anton-Culver H. Medical hypothesis: xenoestrogens as preventable causes of breast cancer. Environ Health Perspect 1993;101:372 – 7.
181
[17] Hoyer AP, Grandjean P, Jorgensen T, Brock JW, Hartvig HB. Organochlorine exposure and risk of breast cancer. Lancet 1998;352:1816 – 20. [18] Moysich KB, Ambrosone CB, Vena JE, Shields PG, Mendola P, Kostyniak P, et al. Environmental organochlorine exposure and postmenopausal breast cancer risk. Cancer Epidemiol Biomarkers Prev 1998;7:181 – 8. [19] Charlier C, Albert A, Herman P, Hamoir E, Gaspard U, Meurisse M, et al. Breast cancer and serum level of organochlorines residues. Occup Environ Med 2003;60:348 – 51. [20] Charlier C, Dubois N, Cucchiaro S, Plomteux G. Analysis of PCBs in human plasma by GC-MS. J Anal Toxicol 2003;27: 42 – 5. [21] Charlier C, Pitance F, Plomteux G. PCBs residues in a breast cancer population. Bull Environ Contam Toxicol 2003;71: 887 – 91. [22] Pauwels A, Covaci A, Weyler J, Delbeke L, Dhont M, De Sutter P, et al. Comparison of POP residues in serum and adipose tissue in a female population in Belgium, 1996 – 1998. Arch Environ Contam Toxicol 2000;39:267 – 70. [23] Gladen BC, Doucet J, Hansen LG. Assessing human PCBs contamination for epidemiologic studies: lessons from patterns of congener concentrations in candians in 1992. Environ Health Perspect 2003;111:437 – 43. [24] Wolff MS, Toniolo PG. Environmental organochlorine exposure as a potential etiologic factor in breast cancer. Environ Health Perspect 1995;103(Suppl. 7):141 – 5. [25] Wolff MS, Camann D, Gammon M. Proposed PCB congener grouping for epidemiologic studies. Environ Health Perspect 1997;105:13 – 4. [26] Oakley GG, Devanaboyina US, Robertson LW. Oxidative DNA damage induced by activation of polychlorinated biphenyls (PCBs): implications for PCB-induced oxidative stress in breast cancer. Chem Res Toxicol 1996;9:1285 – 92. [27] Bucholski KA, Begerov J, Winneke G, Dunemann L. Determination of PCBs and chlorinated pesticides in human body fluids and tissues. J Chromatogr A 1996;754:479 – 85. [28] Holford TR, Zheng T, Maynet ST, Zahm SH, Tessari JD, Boyle P. Joint effects of nine PCBs congeners on breast cancer risk. Int J Epidemiol 2000;29:975 – 82. [29] Garrido-Frenich A, Martinez Vidal JL, Moreno Frias M, Olea Serrano F, Olea N. Quantitative determination of endocrinedisrupting PCBs and organochlorinated pesticides in human serum using GC with ECD and tandem MS. J Mass Spectrom 2000;35:967 – 75. [30] Lucena RA, Allam MF, Costabeber IH, Villarejo ML, Navajas RF. Breast cancer risk factors: PCBs congeners. Eur J Cancer Prev 2001;10:117 – 9. [31] Zheng T, Holford TR, Tessari J, Mayne ST, Owens PH, Ward B, et al. Breast cancer risk associated with congeners of PCBs. Am J Epidemiol 2000;152:50 – 8. [32] Millikan R, De Voto E, Duell EJ, Tse CK, Savitz DA, Beach J, et al. DDE, PCBs and breast cancer among African-American and white women in North Carolina. Cancer Epidemiol Biomarkers Prev 2000;9:1233 – 40.
Polychlorinated biphenyls contamination in women with breast cancer Corinne J. Charlier a,*, Adelin I. Albert b, Liying Zhang b, Nathalie G. Dubois a, Guy J. Plomteux a a
Clinical Toxicology Laboratory, Liege University Hospital, Tour II+5, CHU Sart-Tilman, B 4000 Lie`ge, Belgium b Biostatistics Department, University of Liege, Sart Tilman, B 35, 4000 Liege, Belgium Received 9 March 2004; received in revised form 15 April 2004; accepted 19 April 2004
Abstract Background: Polychlorinated biphenyls (PCBs) are widespread highly resistant pollutants in the environment with potential adverse health effects on humans. The aim of the study was to compare PCBs contamination in women suffering from breast cancer with presumably healthy women. Methods: A gas-chromatography/mass-spectrometry method was used to identify and quantify seven PCBs congeners (IUPAC 28, 52, 101, 118, 138, 153,180) in blood from 60 cases of breast cancer and 60 agematched healthy controls. Results: Cases and controls had similar risk profiles, except for menopausal status (respectively 82% vs. 65%, p = 0.014). PCBs were detectable in 69.1% of the samples. Total blood level of PCBs was significantly different ( p = 0.012) in cases (7.08 F 7.51 ppb) and controls (5.10 F 5.15 ppb). The relationship between PCBs concentrations in serum and risk factor was mainly due to serum levels of PCB153, which were significantly higher in breast cancer women than in disease-free subjects (1.63 F 1.26 ppb vs. 0.63 F 0.78 ppb, p < 0.0001), even after accounting for other potential risk factors. Conclusions: These results suggest that environmental exposure to PCBs may contribute to multifactorial pathogenesis of breast cancer. D 2004 Elsevier B.V. All rights reserved. Keywords: Breast cancer; Endocrine disrupters; PCBs
1. Introduction Polychlorinated biphenyls (PCBs) are complex chemical mixtures that comprise theoretically 209 congeners, some of which are known to cause a wide range of adverse effects on animals and humans [1,2]. PCBs are industrial chemicals that have been used for
* Corresponding author. Tel.: +32-4-3668818; fax: +32-43668889. E-mail address: [email protected] (C.J. Charlier). 0009-8981/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.cccn.2004.04.025
diverse commercial applications such as hydraulic fluids, printing inks, or dielectric fluids for capacitors. The production and use of these compounds were banned in the late 1970s. Together with organochlorine pesticides, PCBs constitute what is called Persistent Organic Pollutants (POPs), due to their great chemical stability, their lipid solubility, and their ubiquitous prevalence in environment [3,4]. Adults are mainly exposed through the consumption of dairy products, meat and fish. The presence of POPs in human serum and adipose tissue has been reported in many studies over the last three decades. Several
178
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
epidemiological studies have suggested that exposure to POPs may increase the risk of endocrine disruption by interference with estrogen metabolism [5 – 7], elevation of the amount of bioavailable estradiol [8], or synergistic effects on estrogen responsive genes [9]. POPs have also been hypothesized to be associated with a variety of effects including hypofertility in man [10], carcinogenesis in man (prostate, testis) [11], immunological perturbations and neurological effects [12], as well as breast cancer in women [13 – 19]. Results concerning breast carcinogenesis are controversial. Despite marked regional differences in prevalent cases (high in North America, Western Europe and Oceania, low in Africa and Asia), the cause of most breast cancers is not yet known. Genetics explains about 10% of the cases. Risk factors for breast cancer include the reproductive life of women (early menarche, nulliparity or late age at first birth, late menopause) as well as hormonal factors, whether endogenous (high levels of estrogens, free or bound to sex hormone binding globulin or exogenous [long term use of hormonal replacement therapy, (HRT)]. The exact role of exposure to other products with implication on the hormonal activity, as may be the case for PCBs, is not yet known. In the present study, we compared the blood levels of seven frequently detected PCBs congeners (IUPAC no. 28, 52, 101, 118, 138, 153 and 180) in women with breast cancer and in non-diseased female controls of the same age. Pollutants were quantified simultaneously using a gas chromatographic analyzer coupled to an ion-trap mass spectrometer detector. The PCBs serum levels observed in cases and controls were compared, while taking into account other risk factors for breast cancer.
2. Material and methods 2.1. Study populations The present study involved 60 women, aged 54.8 F 6.59 years, diagnosed with breast cancer and undergoing a surgical intervention (cases), and 60 control women of the same mean age (54.8 F 6.55 years) selected at random in a population of presumably healthy women consulting for routine systematic cervico-vaginal cytological screening. For each sub-
ject, information regarding age at menarche, pregnancy, breast-feeding, menopausal status and family history of breast cancer (mother, sister or grandmother) were recorded by interview. Postmenopausal women were asked for use of HRT. All patients gave their informed consent for participating in the study. For controls, blood specimens were taken at the time of examination, whereas for cases samples were collected at the time of diagnosis prior to surgery. Blood samples (10 ml) were drawn in the early morning after overnight fasting and were immediately centrifuged with serum specimens kept frozen at 18 jC until assay (within 1 week). 2.2. PCB analysis The quantification of the seven PCBs in serum was done using a gas chromatographic analyser coupled to an ion-trap mass spectrometer detector (Saturn 2000, Varian) by a method described elsewhere [20]. Briefly, calibration samples or serum samples (1 ml) were extracted with 1 ml acetonitrile, 1 ml saturated K2CO3 and 5 ml petroleum ether/diethyl ether (98:2, v/v) after addition of 10 Al of internal standard (aldrin, 10 mg/l). The samples were vortex mixed, then centrifuged (2500 rpm, 5 min) and the supernatant was transferred to a glass collection vial. A repeated extraction with another 5 ml of petroleum ether/ diethyl ether was performed and the supernatant added to the first one. The two fractions were simultaneously transferred to the pre-conditioned extraction column. The solution drawn through the cartridge was 3collected and evaporated under mild nitrogen at 35 jC. The residue was reconstituted in 100 Al hexane and 1 Al was injected in split/splitless (ratio 1:35) mode onto the gas chromatograph. The samples were eluted through a HP5 Trace (30 m, 0.25 mm id, 0.25 Am film) column (Agilent). For detection, the mass spectrometry was operated in the electronic impact mode (70 eV). Selected ion storage was used for identification and quantification. All reagents were of the highest purity available. All PCBs and internal standard (aldrin) were purchased from Dr. Ehrenstorfer (Ausburg, Germany). Samples were analysed in a blind procedure together with controls consisting of samples spiked with 0.5 or 2 ppb of each PCB. Interlaboratory comparison program (AMAP Ring
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
test, Institut National Sante´ Publique, Que´bec) was performed as external quality control.
Table 1 Characteristics of risk factors in women with breast cancer (cases) and in disease-free women (controls) of similar age
2.3. Statistical analysis Serum levels of PCBs were expressed as mean F SD. Correction for lipid content was tested, but since the results were not affected, only crude data are presented. Quantitative continuous results were not categorized but a log-transform was applied to PCBs data to normalize their distributions. All subsequent statistical analyses were carried out on the transformed data. Risk factors and PCBs serum levels in cases and controls were compared by means of ordinary logistic regression, not only for each congener separately but also for all PCBs simultaneously. In every comparison, age was included in the analysis. Results were expressed as odds ratios (OR) and associated 95% confidence intervals (CI). In the multivariate analysis, adjustment was made for potential confounding risk factors: menopausal status, number of full-term pregnancies, lactation, use of HRT and family history of breast cancer. Results were confirmed after backward elimination of not significant effects. All results were considered to be significant at the 5% critical level ( p < 0.05). Calculations were done using the SAS (SAS Institute, version 8.2 for Windows) and S-Plus (version 6.2) statistical packages.
3. Results and discussion 3.1. Risk factors The characteristics of cases and controls are displayed in Table 1. The two groups had similar risk profiles but the prevalence of menopause was significantly higher in women with breast cancer, yielding to an odds ratio of 3.8 (95% CI: 1.3 – 11).
179
Age of menarche, mean years (SD) Menopause, n (%) HRTb, n (%) Parity, n (%) Breast feedingc, n (%) Family history of breast cancer, n (%)
Cases (n = 60)
Controls (n = 60)
pa
11.4 (0.95)
11.5 (1.08)
0.59
49 30 34 24 9
(81.7) (61.2) (56.7) (70.6) (15.0)
39 26 33 25 5
(65.0) (66.7) (55.0) (75.6) (8.30)
0.014 0.45 0.85 0.85 0.25
a
p-value adjusted for age. Restricted to menopaused women. c Restricted to parous women. b
were significantly higher in cases than in controls. The same observation was made for total PCB content ( p = 0.012). 3.3. Association with breast cancer All PCB concentrations and confounding risk factors for breast cancer were combined into a multiple logistic regression analysis (see Table 3). It is seen that menopause ( p = 0.038) and PCB153 ( p < 0.0001) remained statistically discriminate between cases and controls while the effect of PCB138 disappeared. Thus, high concentrations of PCB153 are significantly associated with an increased risk of breast cancer despite the presence of other factors, menopause in particular. After backward elimination of not significant effects, menopause and serum PCB 153 concentration remained significant (respectively, p = 0,036 and p < 0,0001), reinforcing the previous conclusion. Similar results were obtained when replacing individual PCBs by total PCB content. The only two significant variables were menopausal status ( p = 0.015) and total serum PCB concentration ( p = 0.035). 3.4. Discussion
3.2. PCB concentrations PCBs 28 and 118 could not be detected in any of the subjects of the study. The distributions of the five other congeners in cases and controls are presented in Table 2. For PCBs 52, 101 and 180, serum concentrations did not differ in the two groups. By contrast, the serum concentrations of PCB138 and PCB153
In the present case-control study, PCB153, as already suggested by a previous limited study [21] and, to a lesser extent, PCB138 were found in significantly higher concentration in blood samples of women diagnosed with breast cancer when compared to a control group of disease free subjects of the same age. In our study, serum PCB153 concentration
180
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
Table 2 Means (SD) of PCBsa serum levels (ppb) in women with breast cancer (cases) and in disease-free women (controls) of similar age Congener
Cases (n = 60)
Controls (n = 60)
pb
PCB52 PCB101 PCB138 PCB153 PCB180 Total PCB content
1.64 0.77 1.25 1.63 1.79 7.08
1.67 0.62 0.94 0.63 1.24 5.10
0.87 0.23 0.0068 < 0.0001 0.17 0.012
(3.45) (1.21) (1.20) (1.26) (3.46) (7.51)
(3.06) (1.38) (1.84) (0.78) (1.45) (5.15)
a
PCBs 28 and 118 were undetected. p-values were obtained by logistic regression applied to log (PCB) after adjusting for age. b
remained associated with an increased risk of breast cancer, even after accounting for classical breast cancer risk factors. In a previously published paper, the PCB153 level correlated with the sum of other PCBs in blood and was proposed as an indicator substance in monitoring studies [22]. In another recent study [23], congeners 118, 138, 153 and 180 were highly intercorrelated, but not correlated with congeners 28 or 101. Our results are in agreement with preeminence of PCB153 and with correlation between PCB153 and total serum PCB concentration. PCB153 is an ortho-substituted PCB congener, able to present some estrogenic activity such as increasing uterine weight, causing precocious puberty in rats, and enhancing proliferation of MCF-7 breast tumour cells [24]. Moreover, PCBs that have two orthosubstitutions together with two parasubstitutions have been described as phenobarbital-like cytochromes inducers [1,25] and PCB153 could induce oxidative DNA damage implicated in breast carcinogenesis [26]. It is yet unclear what relationship there may be between environmental exposure to these types of compounds and either the initiation or progression of certain diseases. However, these compounds, entering into the human body via the food chain, are characterized by a long half-life and lipophilic properties, which facilitate their accumulation in adipose tissues. They have been detected in human tissues such as blood, milk or fat [27,28]. Blood is the most accessible biological fluid and close correlation between the concentrations in blood and the total body burden can be obtained [29]. Bench top instrumentations have allowed the development of easy methods to initiate epidemiological studies in standardized conditions and
the possible association of PCBs with breast cancer has been postulated [30]. Depending on the reports, absence of correlation [31] but also association with one or another PCB have been reported [28]. In a population of 65 women treated by excision biopsy, PCB28 was found to be the most important risk factor (OR = 9.6) [30]. Association between PCB levels and breast cancer risk was sometimes restricted to subgroups (e.g. obese) of women [32]. In our study, PCB153 appeared to be the ‘‘specific’’ congener if an association with breast cancer risk is evaluated. Moreover, after accounting for potential confounding factors, PCB153 remained a significant risk factor. However, given the complexity of PCB mixtures in environmental exposures, it is reasonable to assume that different spectra of toxicity outcomes may be observed and we agree with other authors [23] that congeners to be measured in any specific epidemiological study will depend on the purpose of the study. In conclusion, our results comfort the debate that there is not sufficient evidence to answer the question on human risk resulting from low-dose endocrine-related effects. Further interdisciplinary research, combining detection and quantitation of pollutants, epidemiological data collection, but also metabolic polymorphism investigations should be conducted to identify individual compounds or pollutants associations that might be contributing to human cancer.
Table 3 Multiple logistic regression analysis applied to risk factors and PCBs concentrationsa in 60 women with breast cancer (cases) and 60 disease-free women (controls) of the same age Risk factors
OR (95%CI)
p
Age (years) 0.94 (0.85 – 1.02) 0.13 Age at menarche (years) 0.89 (0.58 – 1.38) 0.61 Menopause (yes) 4.4 (1.1 – 18) 0.038 HRT(yes)b 0.67 (0.24 – 1.9) 0.45 Parity (yes) 1.5 (0.4 – 5.8) 0.53 Breast feeding (yes)c 0.73 (0.19 – 2.7) 0.64 Family history of breast cancer (yes) 0.84 (0.21 – 3.3) 0.81 PCB52 (ppb) 0.95 (0.74 – 1.2) 0.70 PCB101 (ppb) 1.0 (0.77 – 1.3) 0.95 PCB138 (ppb) 1.2 (0.88 – 1.5) 0.29 PCB153 (ppb) 1.8 (1.4 – 2.5) < 0.0001 PCB180 (ppb) 1.1 (0.76 – 1.5) 0.64 a
A log-transform was used for all PCBs. Restricted to menopaused women. c Restricted to parous women. b
C.J. Charlier et al. / Clinica Chimica Acta 347 (2004) 177–181
References [1] Brouwer A, Longnecker MP, Birnbaum LS, Cogliano J, Kostyniak P, Moore J, et al. Characterization of potential endocrine-related heath effects at low dose levels of exposure to PCBs. Environ Health Perspect 1999;107(Suppl. 4):639 – 49. [2] Mc Kinney JD, Waller CL. PCBs as hormonally active structural analogues. Environ Health Perspect 1994;102:290 – 7. [3] Chevreuil M, Granier L. Les PCBs, des polluants difficiles a` e´liminer. La recherche 1992;23:484 – 6. [4] Safe SH. PCBs: environmental impact, biochemical and toxic responses and implications for risk assessment. Crit Rev Toxicol 1994;24:87 – 149. [5] Korach KS, Sarver P, Chae K, Mc Lachlan JA, Mc Kinney JD. Estrogen receptor binding activity of polychlorinated hydroxy-biphenyls: conformationally restricted structural probes. Mol Pharmacol 1988;33:120 – 6. [6] Robinson AK, Sirbasku DA, Stancel GM. DDT supports the growth of an estrogen-responsive tumour. Toxicol Lett 1985;27:109 – 13. [7] Connor K, Ramamoorthy K, Moore M, Mustain M, Chen I, Safe S, et al. Hydroxylated PCBs as estrogens and antiestrogens: structure – activity relationship. Toxicol Appl Pharmacol 1997;142:160 – 8. [8] Gellert JR. Uterotrophic activity of PCBs and induction of precocious reproductive ageing in neonatally treated female rats. Environ Res 1978;16:123 – 30. [9] Jansen HT, Cookes PS, Porcelli J, Liu T-C, Hansen LG. Estrogenic and antiestrogenic actions of PCBs in female rats: in vitro and in vivo studies. Reprod Toxicol 1993;7: 237 – 48. [10] Oliva A, Spira A, Multigner L. Contribution of environmental factors to the risk of male infertility. Hum Reprod 2001;16: 1768 – 76. [11] Giwercan AE, Skakkebaek NE. The human testis: an organ at risk? Int J Androl 1992;15:373 – 85. [12] Shain W, Bush B, Seegal R. Neurotoxicity of PCBs: structure – activity relationship of individual congeners. Toxicol Appl Pharmacol 1991;111:33 – 42. [13] Cocco P. On the rumors about the silent spring. Review of the scientific evidence linking the occupational and environmental pesticide exposure to endocrine disruption health effects. Cad Sau`de Pu`blica 2002;18:379 – 402. [14] Laden F, Hankinson SE, Wolff MS, Colditz GA, Willett WC, Speizer FE, et al. Plasma organochlorine levels and the risk of breast cancer: an extended follow-up in the nurse’s health study. Int J Cancer 2001;91:568 – 74. [15] Aronson KJ, Miller AB, Woolcott CG, Sterns EE, McGready DR, Lickley LA, et al. Breast adipose tissue concentrations of polychlorinated biphenyls and other organochlorines and breast cancer risk. Cancer Epidemiol Biomarkers Prev 2000;9:55 – 63. [16] Davis DL, Bradlow HL, Wolff M, Woodruff T, Hoel DG, Anton-Culver H. Medical hypothesis: xenoestrogens as preventable causes of breast cancer. Environ Health Perspect 1993;101:372 – 7.
181
[17] Hoyer AP, Grandjean P, Jorgensen T, Brock JW, Hartvig HB. Organochlorine exposure and risk of breast cancer. Lancet 1998;352:1816 – 20. [18] Moysich KB, Ambrosone CB, Vena JE, Shields PG, Mendola P, Kostyniak P, et al. Environmental organochlorine exposure and postmenopausal breast cancer risk. Cancer Epidemiol Biomarkers Prev 1998;7:181 – 8. [19] Charlier C, Albert A, Herman P, Hamoir E, Gaspard U, Meurisse M, et al. Breast cancer and serum level of organochlorines residues. Occup Environ Med 2003;60:348 – 51. [20] Charlier C, Dubois N, Cucchiaro S, Plomteux G. Analysis of PCBs in human plasma by GC-MS. J Anal Toxicol 2003;27: 42 – 5. [21] Charlier C, Pitance F, Plomteux G. PCBs residues in a breast cancer population. Bull Environ Contam Toxicol 2003;71: 887 – 91. [22] Pauwels A, Covaci A, Weyler J, Delbeke L, Dhont M, De Sutter P, et al. Comparison of POP residues in serum and adipose tissue in a female population in Belgium, 1996 – 1998. Arch Environ Contam Toxicol 2000;39:267 – 70. [23] Gladen BC, Doucet J, Hansen LG. Assessing human PCBs contamination for epidemiologic studies: lessons from patterns of congener concentrations in candians in 1992. Environ Health Perspect 2003;111:437 – 43. [24] Wolff MS, Toniolo PG. Environmental organochlorine exposure as a potential etiologic factor in breast cancer. Environ Health Perspect 1995;103(Suppl. 7):141 – 5. [25] Wolff MS, Camann D, Gammon M. Proposed PCB congener grouping for epidemiologic studies. Environ Health Perspect 1997;105:13 – 4. [26] Oakley GG, Devanaboyina US, Robertson LW. Oxidative DNA damage induced by activation of polychlorinated biphenyls (PCBs): implications for PCB-induced oxidative stress in breast cancer. Chem Res Toxicol 1996;9:1285 – 92. [27] Bucholski KA, Begerov J, Winneke G, Dunemann L. Determination of PCBs and chlorinated pesticides in human body fluids and tissues. J Chromatogr A 1996;754:479 – 85. [28] Holford TR, Zheng T, Maynet ST, Zahm SH, Tessari JD, Boyle P. Joint effects of nine PCBs congeners on breast cancer risk. Int J Epidemiol 2000;29:975 – 82. [29] Garrido-Frenich A, Martinez Vidal JL, Moreno Frias M, Olea Serrano F, Olea N. Quantitative determination of endocrinedisrupting PCBs and organochlorinated pesticides in human serum using GC with ECD and tandem MS. J Mass Spectrom 2000;35:967 – 75. [30] Lucena RA, Allam MF, Costabeber IH, Villarejo ML, Navajas RF. Breast cancer risk factors: PCBs congeners. Eur J Cancer Prev 2001;10:117 – 9. [31] Zheng T, Holford TR, Tessari J, Mayne ST, Owens PH, Ward B, et al. Breast cancer risk associated with congeners of PCBs. Am J Epidemiol 2000;152:50 – 8. [32] Millikan R, De Voto E, Duell EJ, Tse CK, Savitz DA, Beach J, et al. DDE, PCBs and breast cancer among African-American and white women in North Carolina. Cancer Epidemiol Biomarkers Prev 2000;9:1233 – 40.