Abstract
Lung cancer is a dominant cause of cancer mortality. The etiology of lung cancer is mainly related to cigarette smoking, airborne genotoxic carcinogens, and arsenic, but its sex-specific incidence suggests that other mechanisms, such as hormones, may also be involved in the process of carcinogenesis. A number of agents commonly present in the living environment can have dual biological effects: not only are they genotoxic / carcinogenic, but they are also hormonally active as xenoestrogens. This dualism may explain sex-specific differences reported in both types and incidence of lung cancer. In a novel approach to investigate the complexity of lung cancer, etiology, including systems biology, will be used as a tool for a simultaneous interpretation of measurable environmental and biological parameters. Using this approach, the etiology of human lung cancer can be more thoroughly investigated using the available data from oncology and environmental health. The information gained could be applied in the introduction of preventive measures, in personalized medicine, and in more relevant legislation, which should be adjusted to reflect the current knowledge on the complex environmental interactions underlying this life-threatening disease.
Introduction
Lung cancer in man is the neoplasm with the greatest incidence and highest mortality rate, and it has, over the past few decades, been mostly associated with cigarette smoking, occupational and environmental exposures to chemicals such as radon and asbestos, and industrial air pollution in general. The current high incidence of lung cancer is thought to be a consequence of lifestyles and environmental exposures that occurred some decades ago, as expected from the relatively long latency period required between exposure and the clinical manifestation of the disease (Besaratina et al. 2008; Sato et al. 2007). It has been estimated that between 3% and 6% of all lung cancer cases have a positive family history of the disease, suggesting a joint role of shared genes and the environment in the development of the disease.
Lung cancer is classified according to pathohistological types into two major groups, small cell lung cancer (SCLC) and non–small cell lung cancer (NSCLC). Non–small cell lung cancer accounts for more than 85% of all lung cancer cases. Non–small cell squamous cell carcinoma (approximately 30%) has been shown to be more frequent in men, whereas adenocarcinoma (approximately 30%–40%) is more frequently diagnosed in women. Adenocarcinoma has been shown to occur more frequently in nonsmokers than in smokers (Clement-Duchene et al. 2010; Singh et al. 2010). Small cell lung cancer, in comparison to NSCLC, grows faster and metastasizes earlier. Interestingly, autopsies in patients with SCLC frequently reveal areas of NSCLC differentiation (Jacot et al. 2001). This finding has significance, in that it suggests that lung carcinogenesis may occur in a mutated, pluripotent stem cell capable of differentiation along several pathways.
The recent advances in molecular biology techniques, and the increased use of molecular biomarkers and genomics, have helped elucidate many new aspects of the etiology of lung cancer and have provided potential new molecular targets in its treatment. It is clear that these new advances in the understanding of lung cancer provide an opportunity for introducing more effective public health policies and a more promising personalization of therapies. To achieve these goals, a clear insight into the complexity of the biological interactions among subjects, between different patient characteristics, between environmental carcinogens / drug interactions including metabolism, and disease pathophysiology, genomic and proteomic technologies, and the integration of the generated data using systems biology and bioinformatics, will be crucial if we are going to adequately make use of the burgeoning data generated through such technologies (Marrer et al. 2006; Schmidt 2006).
In this article, we discuss the interaction, and common features, of selected xenobiotics involved in human lung carcinogenesis, such as air pollution, arsenic, smoking, and indoor radon, and their potential for interaction with endogenous hormones.
Estrogen and Lung Cancer
Recent critical discoveries in oncology (Baik et al. 2004; de Giorgi et al. 2009; Konduri and Schwartz 2007; Spiers and Shaaban 2009; Tanaka et al. 2003) show that the levels of estrogens and progesterone play a significant role in carcinogenesis in multiple organs and tissues. A decreased risk of pulmonary adenocarcinoma in women has been associated with early onset of menopause, related most probably with the decreased levels of estrogen accompanying the menopause stage in a woman’s life (Taioli and Wynder 1994). The levels of circulating estrogen and the expression of estrogen receptors in lung cancer are significant risk factors in both men and women. It has been shown that elevated estrogen receptor β (ERβ) expression is a prognostic factor for survival in men (Schwartz et al. 2005). Although the level of estrogen receptors in NSCLC is not significantly different between men and women, there is a significantly higher expression of estrogen receptors in neoplastic, rather than in nonneoplastic, lung tissue, even in the same patients (Canver et al. 1994; Niikawa et al. 2008). The impact of estrogen level on sex-specific therapy for lung cancer, the achieved specific outcomes, and overall survival are of great current interest (Hede 2007; Mah et al. 2007; Niikawa et al. 2008).
The investigation of common patterns of carcinogenesis in different organs enables a more efficient application of the collected knowledge. An example of this new approach is the identification of possible common genes for lung and breast cancer. Families with breast cancer in women show a significantly higher risk for lung adenocarcinoma in first-generation relatives (Tsuchiya et al. 2007). Additionally, the recently reported observation that specific subtypes of ER receptors found in breast cancer (Herynk et al. 2009; Speirs and Shaaban 2009) may also be present in lung cancer confirms the value of a broader approach in investigating the etiology of cancer in general.
The association between the expression of estrogen receptor and clinico-pathological factors in lung cancer remains unclear. Kawai et al. (2005) reported that ERβ was expressed in the nucleus, and ERα in the cytoplasm, of NSCLC. Nuclear ERβ has also been shown to be present in the cells of the normal lung (Stabile et al. 2002). Both ERα and ERβ have a high affinity for estradiol, and some epidemiologic studies have shown that the expression of ERβ correlates with a favorable prognosis, whereas a negative expression of ERβ predicts poor therapeutic outcome (Schwartz et al. 2005; Skov et al. 2008; Wu et al. 2005).
The expression of epidermal growth factor receptor (EGFR) has a signficant clinical impact on therapy using selective tyrosine kinase inhibitors of EGFR in patients with lung cancer, and overexpression of EGFR is detectable in approximately 80%–85% of patients with NSCLC (Alkis et al. 2010). Specific EGFR mutation in lung cancer is critical in predicting responsiveness to tyrosine kinase inhibition and successful therapy. The specific mutation has been shown to occur more frequently in Asian women, in nonsmokers, and in adenocarcinoma patients than in the general population (Hsieh, Lim, and Kuo 2005; Uramoto, Sugio, and Oyama 2006). Stimulation of ER has also been shown to increase the activity of the EGFR signal, which in turn increases the activity of ER (Pietras and Marquez-Garbab 2007). The association between ER expression and clinico-pathological parameters such as EGFR gene mutation may elucidate the prognostic role of ER expression in adenocarcinoma of the lung. Nose, Sugio, and Oyama (2009) demonstrated that strong nuclear expression of ERβ correlated with the incidence of EGFR mutations and that its favorable prognostic significance is influenced by EGFR mutations in patients with adenocarcinoma of the lung. The additional contribution of environmental and genetic factors involved in lung cancer, such as hormone replacement therapy and oral contraceptives, may also play a relevant role in this response (Schwartz, Prysak, and Murphy 2005; Siegfried 2006).
Xenoestrogens and Lung Cancer
Air Pollution
Air pollution is generally considered to play a significant role in the development of lung cancer (Boldo et al. 2006; Chiu et al. 2006; Cohen 2000; Kabir, Bennett, and Clancy 2007). Air pollution is a complex mixture of particulates (particle pollution or particulate matter, PM) and gas contaminants. Particulate matter is made up of solid and liquid particles suspended in the air and it is composed of a carbon core; acids, including nitrates and sulfates; organic chemicals, including polyaromatic hydrocarbons (PAHs); metals; soil; and dust particles. According to the particles' size, they are categorized into coarse particles (<10 μm and >2.5 μm in aerodynamic diameter, PM10), fine particles (≤2.5 μm and >0.1 in aerodynamic diameter, PM2.5), and ultrafine particles (≤0.1 μm). Although particle size is accounted for in lung cancer risk assessments, very often their composition and their chemical origin are not (Vineis et al. 2004). Combustion of fossil fuels, road traffic, industrial sites, and waste dumps are the major sources of air pollution. Fossil fuel–fired power plants emit a number of different mutagens and carcinogens, such as heavy metals (Pg, Cd, Ni, Cr), PAHs, dioxins, and radionuclides (Parodi et al. 2004), many of which have been shown to possess xenoestrogenic activity. In industrial areas, high levels of PAHs in air correlate with increased levels of DNA adducts in peripheral lymphocytes, and with an increased incidence of lung cancer (Hemminki and Veidebaum 1999). Polyaromatic hydrocarbons are known mutagenic and carcinogenic compounds (IARC 1983) and have estrogen-like activity (Wenger et al. 2008). Lung cancer may be caused by PAHs via still undiscovered pathways in which estrogen interactions may play a key role. Although benzo(a)pyrene is currently used as the main chemical indicator of exposure to mixtures of PAHs, such as dibenzo(a,h)anthracene and dibenzo(a,l)pyrene, appear to be much more carcinogenic (Okona-Mansah et al. 2005) and their estrogen potency is almost unknown.
Sex-specific susceptibility to lung cancer has been detected in air-polluted areas. In an Italian study of high air pollution areas close to coal-fired power stations, high concentrations of heavy metals such as Cd or Ni caused a significant increase in lung cancer mortality in women, but not in men (Parodi et al. 2004). Higher concentrations of lead in the lung tissue of women than in men may result from sex-specific differences in the metabolism and bioaccumulation of lead under the influence of tissue estrogen levels (Fortoul et al. 2005). Carcinogen metabolizing genes have been shown to be regulated by sex hormones in lung cancer tissue, a finding that implies a complex, sex-specific interaction of xenobiotics and estrogen levels (Spivack et al. 2003). Of especial interest are those enzymes involved in estrogen receptor expression, such as quinine oxidoreductase (NQO1), the overexpression of which has been strongly associated with lung cancer (Fowke et al. 2004; Spivac et al. 2003; Vineis and Hoek 2006).
Women non-occupationally exposed to cooking oil fumes have been shown to be at an increased risk of developing lung adenocarcinoma (Zhao et al. 2006). The main substance in cooking oil responsible for lung cancer is thought to be trans, trans-2,4-decadienal, which stimulates the production of reactive oxygen species, cell proliferation, and the release of pro-inflammatory cytokines in the lung and bronchial epithelial cells (Chang, Lo, and Lin 2005; Tung et al. 2001). Exposure to solid fuel combustion products used for cooking or heating has also been associated with increased risks of lung cancer (Lissowska et al. 2005), especially in women (Ramanakumar, Parent, and Siemiatycki 2006).
Arsenic
Arsenic is a known pulmonary carcinogen. Lung cancer caused by arsenic is considered to be a complex interaction of its clastogenic, aneugenic, and hormonally disturbing potency (Kligerman and Tennan 2007). The correlation of arsenic exposure, via drinking water, with lung cancer is more pronounced in older men who habitually drink and smoke; smoking, in particular, seems to show a synergistic effect with arsenic (Chen et al. 2004). The interindividual difference in susceptibility to the genotoxic effects of arsenic is based on polymorphic expression of the enzyme arsenic (III) methyltransferase (Hernandez et al. 2008). Transplacental exposure to arsenic via the drinking water in an animal model in mice showed increased aromatase and ERα mRNA transcription in the fetal lung of female mice. Lung adenocarcinoma in adult female mice transplacentally exposed to arsenic showed higher ERα expression than in lung tissue cells of unexposed animals (Shen et al. 2007; Watson and Yager 2007).
Indoor Radon
Levels of indoor radon can be increased in cases of (a) naturally high radon background levels owing to the soil composition; (b) technologically increased levels as result of surface mining which caused exposure of layers with higher radon levels previously covered by soil layers with low radon levels; and (c) exposure from radon soil exhalation in newly built energy saving dwellings, in which airtight windows were used so as to reduce heating loss (Almgren, Isaksson, and Barregard 2008; Au et al. 1995; Bossew and Lettner 2007; Lavi, Steiner, and Alfassi 2009; Malanca et al. 1992; Somlai et al. 2006; Tung 2004). Exposure can also occur via drinking bottled water, including soft drinks, and in food prepared with contaminated water (Janssen 2003). According to available studies, residential radon is an important cause of lung cancer in the general population, with an excess relative risk of 10% per 100 Bq m3 (Barros-Dios et al. 2002; Catelinois et al. 2006; Kreienbrock et al. 2001; Lagarde et al. 2001; Pershagen et al. 1994). Although no clear difference in histological subtypes of lung cancer induced by exposure to radon has been detected (Field et al. 2001), studies on uranium miners, and in a non-occupationally exposed population, were suggestive of a higher risk for SCLC and adenocarcinoma (Bochicchio et al. 2005; Wichmann et al. 2005) following exposure to this radionuclide. As there appears to be synergism between radon exposure and smoking, exposure to radon is particularly hazardous for smokers and recent ex-smokers (Darby et al. 2005; Catelinois et al. 2006; Lagarde et al. 2001).
Pretreatment with estrogen in an animal model has revealed that α particles, as emitted by radon progeny, cause an increase in the expression of the genes BRCA1, BRCA2, and RAD51 and a cascade of events leading toward the development of breast carcinogenesis (Calaf and Hei 2000). Although no similar study is available for lung carcinogenesis, it would be interesting to investigate whether such mechanisms also exist in lung tissue. A recent clinical study reported a correlation between BRCA1 and the survival of lung cancer patients (Kim et al. 2008). In this study, it was found that coexposures to xenoestrogens (such as those contained in cigarette smoke) and indoor radon in carriers of different BRCA1 polymorphisms and haplotypes, could act synergistically in the process leading to lung cancer development.
Indoor radon protection is, in some regions of the world such as Europe, supported by policies that warn, and financially support, the general population about health risk and procedures that may lead to a reduction the exposure to indoor radon. Policies introduced to provide financial support to homeowners (Janssen 2003) to improve ventilation systems, introduced during the last ten years, have been shown to be extremely cost effective (Coskeran et al. 2006).
Cigarette Smoke
Cigarette smoke is a typical source of complex environmental chemical exposure. More than 3,000 chemicals have been identified in tobacco smoke, and many of them are both mutagenic and carcinogenic (Gram 1997). Cigarette smoke also contains radioisotopes such as 210Po and 210Pb (α particles) obtained from tobacco cultivated in soil contaminated by fertilizers containing high polonium levels. Therefore, cigarette smoking presents a significant exposure of both ionizing and non-ionizing chemical hazard to the lungs (Desideri et al. 2007). Some of the chemicals detected in cigarette smoke have been extensively studied. Nitrosamines have been linked to higher incidence of lung cancers in male than in female animals (Hoffmann et al. 1984), whereas higher levels of polycyclic aromatic hydrocarbon–derived DNA adducts have been reported in female smokers than in male smokers. Even though this finding may be explained by a higher expression of CYP1A1 in women, there are indications that even nonsmoking women are more susceptible to lung cancer than nonsmoking men (Schwartz , Prysak, and Murphy 2005). These findings suggest that the female sex has a higher inherent risk, probably through their greater expression of estrogen and its associated receptors, of developing lung cancer from agents other than those found in cigarette smoke (Ramchandran and Patel 2009).
Transplacental exposure studies showed that gene expression in the offspring may be modified by exposure of the mother to tobacco smoke, probably through an alteration in methylation levels of the DNA, with health consequences expressed during the subsequent adulthood (Rapiti et al. 1999; Rouse, Boudreaux, and Penn 2007).
Since nicotine and cotinine are aromatase inhibitors, cigarette smoke has also been shown to decrease estrogen levels in smokers (Barbieri, Gochberb, and Ryan 1986), although whether this decrease is directly associated with two chemicals rather than to the cocktail present in the smoke remains debatable. Inhibition of aromatase leads to increased testosterone levels in female smokers (Sowers et al. 2001), which could be related to higher androgen expression in NSCLC (Nishio et al. 2005).
The synergistic activity of smoking with exposure to xenobiotics, such as arsenic, asbestos, and radon, in cancer risk is well known (Ahsan and Thomas 2004; Arnold and McLachlan 1996; Hassan et al. 2007; Hays et al. 2006; IARC 2004; Lee et al. 2009). The mechanism is thought to operate through the CYP enzymes, especially CYP1A2, which takes part in the activation of carcinogens (Nishikawa et al. 2004).
In a study of the intake of red and processed meat that contains heterocyclic amines and benzo(a)pyrene, compounds that bind to estrogen receptors, a positive association with lung cancer incidence was detected. This association was even more pronounced among never smokers than among smokers (Bennion et al. 2005; Grover et al. 1980; Lam et al. 2009).
An alcohol-induced increase of estrogen synthesis has been reported among smokers (Coutelle et al, 2004; Fan et al. 2000), and a synergistic effect of alcohol and smoking has been described for squamous cell carcinoma of the esophagus (Hashibe et al. 2007), an estrogen-positive cancer (Kalayarasan et al. 2008).
Conclusion
Current knowledge of environmental health, oncology, and epidemiology gives new insight into the etiology of lung cancer as a consequence of the interaction between the gene pool and the complex chemical environment of the present time. Models for the risk assessment of lung cancer, which is clearly a multifactorial disease, are still based on reductionism, in which etiology is interpreted on single-agent causality of exposure and disease. Available mathematical solutions are still not applied in the evaluation of the complex interactions between the biology of humans and/or their chemical exposure, and any consequent adverse health effects. Failure to adopt a truly multifactorial analysis of the available data may give the false impression that some types of lung cancers appear only by chance, and such an approach may miss the possibility of a relationship with exposure to multiple hazardous xenobiotics. A crucial step forward in understanding the etiology of lung cancer would be a risk assessment in which the role of several agents, known to be present in the environment, are accounted for together with the role played by known confounders or effect modifiers, using artificial intelligence software.
Future studies on etiology of lung cancer should synthesize (a) exposure data, (b) investigation of the association of genetic polymorphisms in estrogen metabolism, (c) ERβ receptor polymorphisms and their isoform expression (Swartz et al. 2005), and (d) results of biomonitoring data collected in environmental health. Lung cancer types, with etiology related to environmental exposure, could be correlated with exposure to xenobiotics that express xenoestrogenic or aromatase inhibitor activity. Closer collaboration between oncology, systems biology, and environmental health may in future provide a significant qualitative leap forward in the elucidation of lung cancer etiology using existing knowledge. Results of such collaborations will be significant for improving the living environment through more relevant legislation that incorporates analysis of all of the data in a clearly understood fashion, allowing modeling of therapy on an individual basis and the potential for developing new diagnostic biomarkers and drugs.
