Nontobacco-Related Lung Carcinogenesis
Tze-Ming Benson Chen
Ware G. Kuschner
Lung cancer is estimated to cause 17,000 to 26,000 deaths among nonsmokers annually in the United States.1 Secondhand smoke exposure explains some deaths among nonsmokers, but many deaths are unrelated to tobacco smoke. Occupational and environmental exposures and genetic characteristics have been identified as risk factors for the development of lung cancer in both smokers and nonsmokers. In this chapter, we review important toxicants, other than tobacco, for which the evidence for pulmonary carcinogenic potential is strong and the population health effects data support a causal relationship between exposure and lung cancer. These toxicants share the common feature of being respirable carcinogens, but otherwise have widely different physicochemical characteristics. The substances include minerals (asbestos and silica), radioactive gas (radon), and products of fossil fuel combustion (diesel-exhaust particles).
Historically, occupational settings have been among the most important sources of exposure to the nontobacco pulmonary carcinogens. Accordingly, many cases of lung cancer attributable to nontobacco carcinogens are work related and, in turn, preventable. Moreover, we review data on some exposures mainly encountered in industrial settings, which are most persuasively linked with lung cancer. It is important to note, however, that since the beginning of the 20th century, an estimated 85,000 chemicals have been introduced into industrial applications, many of which may be encountered in respirable states.2 Data on the carcinogenic potential of most of these chemicals are limited or nonexistent. Some cases of lung cancer may be attributable to occupational or environmental exposures yet to be recognized as carcinogenic. However, such speculation should not alter the fact that strategies to reduce the total burden of lung cancer worldwide must remain sharply focused on preventing exposure to established pulmonary carcinogens, including tobacco and nontobacco exposures.
ARSENIC
Arsenic is a naturally occurring element found throughout the earth’s crust. Inorganic arsenic complexes are used predominantly to preserve wood, whereas organic arsenic is used in pesticides. In addition, arsenic trioxide (arsenite) is used in the treatment of promyelocytic leukemia.
Concerns over the potential carcinogenicity of arsenic were first raised as early as 1820 when Paris3 first described its association with skin cancer. In 1930, Saupe4 described two cases of lung cancer in association with arsenic exposure. Since then, a large amount of data has been published linking arsenic and lung cancer in humans. Blot and Fraumeni Jr5 uncovered an increased risk of lung cancer-related mortality in smelter workers exposed to arsenic trioxide between 1938 and 1963. Tokudome and Kuratsune6 found a significantly increased mortality rate from lung cancer in copper smelters employed at a metal refinery in Japan between 1949 and 1971. The average latency period for lung cancer was 37.6 years and was unrelated to the estimated levels of arsenic exposure. Rencher et al.6a conducted a retrospective mortality study at a copper smelter in Utah and demonstrated that 7% of all worker deaths were caused by lung cancer compared to 2.7% in the state of Utah and 2.2% at the smelter’s associated mine and concentrator. Other investigators have confirmed the association between lung cancer and working directly with copper smelters.7,8,9,10,11 Other studies have found increased risks of lung cancer in association with exposure to pesticides containing inorganic arsenic12,13 and even the use of arsenic as a medicinal.14 Of note, Guo et al.15 concluded arsenic exposure is most strongly associated with the development of either squamous cell or small cell lung cancer.
Most studies assumed that only inhalational exposures are associated with arsenic-related lung cancer. However, along the southwest and northeast coasts of Taiwan, in the Niigata Prefecture of Japan, in Northern Chile, and in Bangladesh, ground water is heavily contaminated by arsenic, and its ingestion has been associated with increased incidences of lung cancer. 16,17,18,19,20 Chen et al.21 evaluated the dose-response relationship between ingested arsenic and lung cancer risk as it relates to smoking. They found an increase in the relative risk of lung cancer of 3.29 (95% confidence interval [CI], 1.60, 2.78) among populations exposed to the highest (700 μg/L) relative to the
lowest (<10 μg/L) arsenic levels in drinking water. In addition, after the instillation of a tap-water system in southwestern Taiwan, lung cancer mortality declined, further bolstering the likely relationship between ingested arsenic and lung cancer.22
lowest (<10 μg/L) arsenic levels in drinking water. In addition, after the instillation of a tap-water system in southwestern Taiwan, lung cancer mortality declined, further bolstering the likely relationship between ingested arsenic and lung cancer.22
The most significant confounding factor in these studies of arsenic and lung cancer is the contribution of smoking. Some more recent studies have tried to adjust for the effect of smoking and have demonstrated a persistent carcinogenic effect of arsenic.23 In addition, a synergistic interaction between arsenic and smoking likely exists as suggested by Pershagen et al.24 This study evaluated Swedish copper smelter workers and found that the age-standardized rate ratio for lung cancer death in arsenicexposed nonsmokers was 3.0. Among those smokers without occupational arsenic exposure, the ratio was 4.9. In arsenic-exposed smokers, the age-standardized rate ratio for lung cancer was 14.6.24 A metaanalysis by Hertz-Picciotto et al.25 and the previously described study by Chen et al.21 also support a synergistic effect.
Nonoccupational exposure to copper smelters (i.e., residential proximity) may also pose a carcinogenic risk. However, most studies failed to establish a statistically significant link within this setting suggesting that arsenic alone may either be a weak carcinogen or may require a cocarcinogen to induce the development of cancer.26,27,28
Reviews of the literature suggest that the average latency for lung cancer diagnosis after exposure to arsenic is about 30 years. In addition, arsenic-related pulmonary malignancies appear to have a predilection for the upper lobes.29 All histologic cell types are represented in arsenic-related lung cancer and the relative frequencies of each cell type seem to mimic that of the general nonexposed population.30,31
Animal data supporting a carcinogenic role for arsenic are limited. Ishinishi et al.32 found that intratracheal instillation of three forms of arsenic (copper ore, flue dust, and arsenic trioxide) to Wistar-King rats was associated with the formation of lung adenomas and/or adenocarcinomas. In another publication, Ishinishi et al.33 demonstrated a 10% to 30% lifetime risk of lung adenocarcinoma in Syrian golden hamsters after weekly intratracheal instillation of 3.75 or 5.25 mg of arsenic trioxide. Ivankovic et al.34 demonstrated the induction of multifocal bronchogenic adenocarcinomas and bronchoalveolar cell carcinomas in 9 of 15 (60%) rats after intratracheal instillation of 0.1 mL of a vineyard pesticide containing calcium arsenate. Soucy et al.35 found dose-dependent effects of arsenic trioxide on animal models of angiogenesis, as well as melanoma tumor growth and metastasis. Interestingly, the form of arsenic appears to influence the risk of lung cancer in animal studies. Specifically, calcium arsenate appears to have the strongest tumorigenic potential, whereas arsenic trioxide is of questionable carcinogenicity.36,37
Arsenic has been shown to induce preneoplastic changes in human fetal lung tissue.38 The mechanism behind such changes may lie in arsenic-induced overmethylation of DNA. Mass and Wang39 found that exposure of human lung adenocarcinoma A549 cells to sodium arsenite or sodium arsenate resulted in a significant level of methylation of a fragment of p53, a tumor suppressor gene. This may alter the function of p53 as a checkpoint in the cell cycle permitting eventual transformation into an immortal cell line. Other mouse studies have suggested that arsenic augments the ability of the tobacco-derived carcinogen, benzo(a)pyrene, to increase the number of DNA adducts in both skin and lung, the initiation step in mutagenesis.40
In 1980, the International Agency for Research on Cancer (IARC) concluded that the available human data were sufficient to implicate arsenic as a pulmonary carcinogen. In October 2001, the U.S. Environmental Protection Agency (EPA) announced that on January 2006, a permissible exposure limit (PEL) of 10 ppb in drinking water would be enforced.41
The National Institute for Occupational Safety and Health (NIOSH) has established a PEL of 2 μg/m3 during a 15-minute ceiling, whereas the Occupational Safety and Health Administration (OSHA) has established a PEL of 10 μg/m3 during any 8-hour period for a 40-hour workweek. Potential household exposures to arsenic through ant pesticides containing sodium arsenate and arsenic-treated pressurized wood prompted the EPA to begin phasing out these products in 1989.
ASBESTOS
Asbestos, derived from a Greek adjective meaning inextinguishable or unquenchable, is a naturally occurring mineral used widely in the 20th century for its insulating and corrosion-and fire-resistant properties. In the late 1800s, the British discovered that asbestos fibers could be woven into textiles permitting its use in everything from brake pads to ship boiler insulators.42 Asbestos had already been in use for centuries, and its associated adverse health effects had been recognized since at least the time of the Roman Empire when Pliny the Elder, a Roman citizen, noticed that slaves working in asbestos mines succumbed early to lung diseases. It was not until the United Kingdom Annual Report of the Chief Inspector of Factories in 1898 that the potential deleterious effects of asbestos were recognized again.43
Doll44 published the landmark epidemiologic study linking lung cancer and asbestos exposure when he evaluated the autopsy results of 105 employees of an asbestos factory. Selikoff45 provided further supportive epidemiologic evidence after reviewing the medical records of 1522 members of the asbestos workers unions in New York City and New Jersey. Wagner et al.46 and Newhouse et al.47 determined that even casual nonoccupational exposure to asbestos was sufficient to cause lung cancer by recognizing epidemics of mesothelioma among communities surrounding asbestos mines and neighborhoods located near asbestos textile mills.
Recent reports suggest that the mutagenic effect of asbestos involves proto-oncogenes, such as k-ras48 and c-ras,49 as well as tumor suppressor genes, such as p53. Nelson et al.50 found a fivefold increase in the presence of k-ras mutations in patients diagnosed with lung adenocarcinoma who had occupational asbestos-exposure history compared to those patients with lung
cancer without exposure history. Panduri et al.51 found that p53 induces alveolar epithelial cell apoptosis in cells damaged by asbestos exposure. Supporting the important protective role of p53, Morris et al.52 demonstrated a fivefold increase in the incidence of asbestos-associated lung cancer in mice after disrupting intrinsic p53 function. Additional findings include alterations in the insulin receptor pathway and associated downregulation of deleted in colorectal cancer (DCC) gene, KU70, and heat shock protein 27.53 The effects of these gene alterations are the constitutive expression of proteins promoting cell division and the downregulation or removal of proteins involved in checkpoints during the cell cycle.
cancer without exposure history. Panduri et al.51 found that p53 induces alveolar epithelial cell apoptosis in cells damaged by asbestos exposure. Supporting the important protective role of p53, Morris et al.52 demonstrated a fivefold increase in the incidence of asbestos-associated lung cancer in mice after disrupting intrinsic p53 function. Additional findings include alterations in the insulin receptor pathway and associated downregulation of deleted in colorectal cancer (DCC) gene, KU70, and heat shock protein 27.53 The effects of these gene alterations are the constitutive expression of proteins promoting cell division and the downregulation or removal of proteins involved in checkpoints during the cell cycle.
Possible mechanisms by which asbestos damages DNA appear to involve the production of reactive oxygen species and the activation of mitogen-activated protein kinases. Iwata et al.54 detected the generation of reactive oxygen species by polymorphonuclear lymphocytes after exposure to anthophyllite asbestos fibers. Schabath et al.55 demonstrated that homozygotes for the G-myeloperoxidase allele (G/G) exhibited an increased risk of asbestos-related lung cancer (odds ratio [OR] 1.72, 95% CI, 1.09 to 2.66) compared to those subjects with G/A and A/A genotypes. Another possible mechanism proposed by MacCorkle et al.56 involves the interaction of asbestos fibers with cell’s cytoskeletal proteins or proteins involved in cell division resulting in an increase in aneuploid cells.
Another theory regarding the carcinogenicity of asbestos hypothesizes that the asbestos fiber’s role is to facilitate the introduction of other carcinogens like those in cigarette smoke to cells. The fibers do so by adhering to surfactant, which then creates a lipid bilayer permitting solubilization of hydrophobic carcinogens such as polycyclic hydrocarbons. This would then permit long-term high concentration exposure of the lung epithelium to carcinogenic substances.57,58,59
The latency period for the development of asbestos-related lung cancer is in excess of 20 years.60 Asbestos has been linked to all cell types of lung cancer.61 The risk of lung cancer in persons exposed to asbestos seems to depend on the fiber type (greater with nonchrysotile fibers even though chrysotile exposure is associated with lung cancer),62 fiber size (greater with longer fibers),63 exposure environment (greater in textile than in cement industries), and evidence of asbestosis on chest radiograph (greater in patients with opacities).64,65,66 The estimated risk of lung cancer in some studies is about fivefold compared to the nonexposed general population. Smoking acts synergistically with asbestos and increases the risk of lung cancer almost 50-fold.45,67
The Asbestos Regulations of 1931 were the first attempt to regulate asbestos exposure in the workplace. Unfortunately, the permitted exposure levels were based on a study of workers at a North Carolina asbestos factory, all of whom had been employed for less than 10 years. Another methodologic problem occurred when about 150 workers were fired prior to the initiation of the study because of concerns that they may have had asbestosis. The Asbestos Regulations of 1969 decreased the permitted exposure level 15-fold to 2 fibers/mL of air. However, a company physician untrained in epidemiology based this “safe” level of exposure on an industry-sponsored study. In 1994, the United States lowered the “safe” exposure level to 0.1 fibers/mL of air, whereas Great Britain banned the use of the substance altogether in 1999. In 2001, the World Trade Organization stated that no safe level of asbestos exposure existed.68 In 1973, the IARC concluded that asbestos was a human lung carcinogen.69
As of year 2000, based on data obtained from death certificates in the United Kingdom, deaths from asbestos-related lung cancer and mesothelioma continue to rise. The implication is that those exposed to the substance over the preceding 20 to 40 years will continue to be at risk for developing lung and/or pleural cancers despite cessation of exposure.42
Recognition that asbestos is the underlying etiology of a patient’s illness should prompt physicians to make the appropriate notifications. In addition, it would be prudent to instruct the patient on the importance of avoiding further exposure to asbestos preferably by changing jobs or using protective respiratory equipment.
Asbestos-Related Mesothelioma Asbestos is the predominant cause of mesothelioma worldwide. In about two thirds of cases, an asbestos-exposure history is present. The risk of mesothelioma varies with the duration and intensity of exposure, as well as the type of asbestos fiber inhaled (highest with amosite and crocidolite). The latency period for mesothelioma is at least 25 to 30 years, and there have been reports of cases occurring more than 40 years after exposure.70 Diagnosis often requires open-lung biopsy and unfortunately, mesotheliomas are notorious for growing along needle tracts and through surgical incisions. The most important step in evaluating a possible mesothelioma involves distinguishing it from benign mesothelioma, primary bronchogenic adenocarcinomas, and metastatic disease given the potentially different treatment options and outcomes.
BERYLLIUM
Beryllium is a naturally occurring element found in soil, rocks, coal, and oil. It was first discovered more than 2 centuries ago but was not widely used in industry until the 1940s and 1950s. Beryllium can withstand extreme heat, remain stable over a wide range of temperatures, and act as an excellent thermal conductor. It also enhances other metals when combined with them as alloys. It is essential for numerous items used in our day-to-day activities. Electrical connections in our cell phones, battery contacts, high-definition and cable television, power steering, electronic ignition systems, and air bag sensors are all modern-day systems and/or appliances that rely on beryllium to function. Mancuso et al.71,72,73,74,75,76 first reported a potential link between beryllium and human lung cancer followed by other reports. However, other reports found no such relationship.77,78,79,80,81,82
Early animal studies suggested the possible role of beryllium in lung cancer; however, the mechanism is still unknown. 83 Studies have attempted to demonstrate a genotoxic
event but have been mostly unsuccessful.80,81,82 One study suggests a potential role for overmethylation of p16, a tumor suppressor gene; however, further work is needed to clarify its possible contribution.84
event but have been mostly unsuccessful.80,81,82 One study suggests a potential role for overmethylation of p16, a tumor suppressor gene; however, further work is needed to clarify its possible contribution.84
The varying results of beryllium studies have prompted four IARC meetings regarding the element’s classification as a pulmonary carcinogen. In 1993, the working group of the IARC concluded that the evidence in human studies is sufficient to implicate beryllium as a carcinogen. The EPA has established that industries may release a total of 0.01 μg of beryllium per cubic meter averaged over 30 days. The OSHA has set a PEL of 2 μg of beryllium per cubic meter of air over an 8-hour workday.85
CHLOROMETHYL ETHER AND BIS(CHLOROMETHYL) ETHER
Used predominantly in industries to synthesize plastics, organic chemicals, and exchange resins, chloromethyl methyl ether (CMME) and an associated impurity, bis(chloromethyl) ether (BCME) were initially produced and used in this country soon after the end of World War II. Their potential role as carcinogen was not realized until 1968 when van Duuren86 demonstrated the development of skin cancers in mice after exposing them to CMME. Leong et al.87 found that BCME and CMME were pulmonary carcinogens after exposing A/Heston mice to vapors of each chemical 6 hours a day, 5 days a week, for a total of 82 to 130 exposure days. Laskin et al.88 confirmed this inhalation effect of BCME on rats and hamsters.
The results of these initial animal studies prompted further human investigation. Albert et al.89 evaluated lung cancer mortality at six of the seven U.S. companies that used CMME at that time. They found a 2.5-fold increase in lung cancer-related mortality among workers exposed to CMME choosing nonexposed employees at those same plants as controls. They also revealed an increase in lung cancer risk with increasing duration and intensity of exposure. DeFonso et al.90 found that CMME with 0.5% to 4% BCME resulted in a 3.8-fold increase in lung cancer risk at a Philadelphia chemical plant, again using employees without an exposure history from the same plant as controls. Numerous studies have provided additional evidence supporting CMME and BCME’s role as pulmonary carcinogens.91,92,93,94,95,96,97,98,99
The literature suggests that the latency period for lung cancer after exposure to BCME and CMME is between 21 to 25 years and is inversely related to exposure intensity and duration. A predominance of small cell carcinomas (80% to 90% of cases) is apparent in most studies.
In the 1970s, restrictions in the use of and safer handling techniques for CMME and BCME were instituted with an apparent decline in the incidence of associated cases of lung cancer.99 Currently, the use of these substances is highly restricted, thereby minimizing potential exposure. BCME breaks down easily and quickly when exposed to sunlight or water and fortunately, does not build up in the food chain. Consequently, the only likely sources of exposure today are living near and/or working in industries that still use BCME and CMME.
The IARC officially recognized CMME and BCME as carcinogens in 1987. At this time, the EPA has set a tolerable limit of 0.0000038 parts of BCME per billion parts of water (0.0000038 ppb) in lakes and streams. Any release of more than 10 lb of BCME into the environment must be reported to the EPA. The OSHA has mandated that no more than 1 ppb of BCME be present in the air of a work environment.100
CHROMIUM
Chromium is a naturally occurring mineral found throughout the environment. It is present in multiple forms and is used in wood preserving, dyes, chrome plating, leather tanning, and steel production.
In the United States, recognition of chromium as a potential pulmonary carcinogen began when the chromate industry acknowledged a concern over the incidence of lung cancer among their employees. This prompted Machle and Gregorius101 to perform a retrospective mortality study encompassing the years 1933 through 1946 of employees at seven different plants located in New Jersey, New York, Maryland, and Ohio. They compared their findings with mortality data for industrial policyholders of the Metropolitan Life Insurance Company for the first 10 months of 1947. They found a 16-fold increase in the risk of lung cancer mortality among employees (range 18 to 50). Several ensuing studies have confirmed the increased risk of lung cancer associated with chromium exposure in occupations ranging from chromate production to the use of chromate-based pigments and/or spray paints.102,103,104,105,106,107,108,109,110,111,112,113 Masonry,114,115 hard-chrome plating,116,117,118 and stainless steel production119 are other occupations associated with chromium exposure and increased risks of lung cancer.