Tobacco carcinogenesis has remained a focus of research during the past 10 years, and various epidemiological and experimental studies have not only confirmed the major role of tobacco smoke exposure in lung and bladder cancers, but have also reported on its association with cancers of various other sites, such as the oral cavity, esophagus, colon, pancreas, breast, larynx and kidney. It is also associated with leukemia, especially acute myeloid leukemia.
In addition to the highly recognized role of cigarette smoking in lung cancer, it has been implicated in many other chronic diseases, including chronic bronchitis and pulmonary emphysema. In the United States, the reduction in smoking has resulted in a decline in death due to lung cancer in men since the mid 1980s. However, the incidence of lung cancer in women has surpassed that of breast cancer and continues to rise; it will likely be the focus of future studies (29
). Both active and passive smoking are implicated in this increase, and several studies of smoking behaviour and disease incidence in women suggest greater susceptibility of women to tobacco carcinogens (31
). It is believed that 80% to 90% of all respiratory cancers are related to active smoking.
Because of the antiestrogenic protective effects of smoking, the role of smoking in breast cancer is controversial. However, recent studies suggest that both active and passive smoking may have a role in the occurrence of breast cancer. One example is a study that found an OR of 4.5 for breast cancer among women who were exposed to passive smoke before 12 years of age and an OR of 7.5 for active smokers. Women who were first exposed to passive smoke after 12 years of age had a lower, although still elevated, OR (32
In both men and women, cancers of the head and neck are also on the rise, and this has been attributed to increased use of smokeless tobacco products. Also, a synergistic interaction between cigarette smoking and radon exposure was confirmed in a large study that showed that lung cancer incidence due to an interaction between smoking and radon exposure exceeded incidence accounted for by additive effects and, therefore, indicated multiplicative effects (33
Comparative toxicity studies have shown that in comparison with standard cigarettes, the new experimental cigarettes that heat tobacco have a relatively low toxicity (34
). In comparing lung cancer risk in smokers of different types of cigarettes, Lee (35
) determined in 2001 that the risk was 36% lower in individuals smoking filtered cigarettes than in those smoking unfiltered cigarettes, and the risk was 23% lower for smokers of low-tar cigarettes than smokers of high-tar cigarettes. The risk increased by 42% in hand-rolled cigarette smokers and by 75% in smokers using black tobacco.
One interesting observation relates to the nature of lung cancer, which has changed over the years with respect to the location and the types of lung tumours observed in smokers. In the past, the primary tumours observed among smokers were the centrally located squamous cell carcinomas of the airways. Now, the predominant lung tumours in smokers are peripheral adenocarcinomas and other non-small-cell lung cancers. This shift in tumour types has been attributed to changes in the composition of cigarettes and its effect on the smoking patterns of tobacco users over the past 30 years (8
). Significant reductions in cigarette tar and nicotine and increased levels of nitrates in cigarettes have markedly altered the manner in which cigarettes are smoked. The number and volume of puffs taken by smokers have increased from a single 35 mL puff/min with 1950s cigarettes to two to four 50 mL puffs/min of low-tar or low-nicotine cigarettes; the depth of inhalation has also increased. These changes in smoking patterns have promoted greater deposition of smoke constituents into the peripheral lungs, where adenocarcinomas develop.
Major advances are being made in the area of molecular epidemiology of tobacco-related cancers in human populations. Many recent epidemiological studies have focused on the differential susceptibility to tobacco-related cancers; they have employed polymerase chain reaction-based molecular assays that permit genotypic analysis of small human samples and supplement the information generated by enzymatic and immunological assays. These assays are increasingly being used in human and experimental studies to examine various gene-gene and gene-environment interactions. One area that has received considerable attention in recent years is the role of polymorphic enzymes in the development of diseases. It is now well recognized that genetic polymorphism strongly influences cancer susceptibility and incidence. The frequencies of mutated alleles of proto-oncogenes, tumour suppressor genes and xenobiotic bio-transformation genes vary significantly among different populations and impact substantially on their susceptibility to cancer. Nearly every enzyme in the carcinogen metabolism pathways has been found to exist in multiple forms, many of which vary in binding affinity and/or turnover efficiency. Some are even entirely absent in individuals, thereby influencing their susceptibility to disease development.
The chemical complexity of tobacco smoke and the metabolic activation requirements for many of its carcinogenic constituents have drawn particular attention to genetic polymorphisms of biotransformation enzymes that metabolize tobacco smoke carcinogens. Thus, genes for various activating enzymes such as cytochrome P450 (CYP) proteins, and deactivating enzymes such as glutathione S-transferase (GST), N-acetyl transferase (NAT) and uridine diphosphate-glucose transferase have been the main target of many recent studies in the context of tobacco carcinogenesis. Also, pre-existing inherited mutations and/or mutation susceptibility of tumour suppressor genes such as p53
, which are known to play a major role in determining cancer susceptibility, have been a subject of investigations in tobacco-related carcinogenesis (37
Several human studies have suggested a strong interplay of various polymorphic CYP1A1, CYP1A2, CYP2E1, NAT1, NAT2, GSTM1 and GSTT1 enzymes in modulating the formation of DNA adducts, induction of mutations and chromosomal damage, and/or the incidence of cancers of various sites in different populations (39
gene has been extensively studied in Japanese populations. Two polymorphic variants that interact with smoking to modify lung cancer risk have been identified (47
). Thus, a homozygous minor allele combined with smoking was found to increase lung cancer risk. Studies of the same gene in Western populations have, however, yielded negative or conflicting results (49
), although an interaction of CYP1A1
variants with the GST null genotype has been reported to significantly increase lung cancer risks in non-Japanese populations (50
NATs are polymorphic conjugation enzymes (produced by the NAT1
genes) involved in the detoxification of aromatic amines by N-acetylation. Depending on the presence or absence of a particular variant, individuals can be categorized as slow or fast acetylators, which in turn can influence the incidence of bladder cancer. It was shown that slow acetylator NAT2 is an important modifier of the amount of aromatic amine-DNA adduct formation even at a low dose of tobacco smoke exposure (52
). Slow acetylator NAT2 genotype was also a significant risk factor for bladder cancer in moderate and heavy smokers, but had no effect in nonsmokers (53
GSTs are another group of metabolic detoxification enzymes that have attracted a great deal of interest in recent years because of their association with risks for different types of cancers. Based on their sequences, these enzymes are divided into five classes. Three of these classes – GSTM1, GSTT1 and GSTPi – are important in the context of tobacco-related cancers. Extensive studies on the relationship of these genes to cancer risks have shown that most populations studied have very high frequencies (20% to 50%) of homozygous GSTM1 and GSTT1 deletion carriers. GSTM1 and GSTT1 may be involved in the etiology of cancer at more than one site. Furthermore, the risk to individuals who carry homozygous deletions is generally small but increases significantly on interaction with cigarette smoking (54
). Among all metabolic cancer susceptibility genes, the association of GSTM1 deficiency with cancer risk is the most consistent and unidirectional. Various experimental and epidemiological observations support the role of this gene in tobacco-related cancers. For example, it has been observed that the excretion of urinary mutagens and the number of lung tissue DNA adducts in GSTM1-deficient smokers is significantly greater than those carrying the wild-type allele (55
). Various epidemiological studies also support the premise that deficiency of this enzyme predisposes for lung and bladder cancers (58
). Furthermore, low activity alleles of GSTPi have been often found in association with different types of human cancers (59
In addition to anomalies of biotransformation enzyme genes, inactivation of tumour suppressor genes such as p53
, and activation of the proto-oncogene K-ras
are also involved in tobacco-related cancers. Various mutated forms of tumour suppressor gene p53
have been commonly detected in lung tumours and it has been found that these mutations are predominantly located in exons 5 to 8. The nature of point mutations in this gene has been extensively investigated and studies show that the most common mutant allele of the p53
gene possesses a G:C to A:C transversion (61
), which is associated with tobacco use (62
The above studies show that several genetically controlled polymorphic enzymes and enzyme systems are linked to tobacco carcinogen activation and deactivation. Some of these genes have been identified and characterized, but others remain undiscovered. Not only the independent effects of single gene polymorphisms, but an interplay of multiple gene interactions appear to be involved. The complexity of epidemiological studies, which have many uncontrollable variables, makes it difficult to study such interactions and their control in human studies. Additionally, many of the enzymes involved in tobacco carcinogen metabolism are also induced by other environmental factors such as alcohol use, dietary constituents, pesticide and xenobiotic exposure, hormonal status, etc, further complicating the interpretation of data. The interaction of many of these genes with each other and the effect of environmental factors are just beginning to be examined. Experimental studies in specifically constructed transgenic and knock-out animals will be important for a systematic evaluation of the contribution of specific cancer genes and/or cancer susceptibility genes to the tobacco carcinogenic process, and to help identify the mechanisms through which environmental agents, such as cigarette smoke, influence these processes.