Epidemiology and Etiology of Lung Cancer | 1 |
INCIDENCE AND MORTALITY
In 2015, lung cancer was the second most common cancer diagnosed in both men and women in the United States, accounting for 14% of all new cancer diagnoses among men and 13% among women. However, lung cancer continues to be the leading cause of cancer-related mortality, accounting for 28% of cancer deaths in men and 26% in women. In 2015, there were 221,000 new diagnoses and 158,000 deaths due to lung cancer in the United States (1). The median age at diagnosis is 70 years, with little difference by sex. In men, incidence and mortality rates increased steadily until the early 1980s and early 1990s, respectively, following a decline in cigarette smoking. In women, incidence and mortality rates did not decrease until the mid-2000s. Average annual age-adjusted incidence rates (2008–2012) are 70.1 per 100,000 men (down by ~30% from peak rates) and 50.2 per 100,000 women (down by ~5% from peak rates); the respective mortality rates are 59.8/100,000 and 37.8/100,000 in men and women (2).
In the United States, incidence and mortality rates vary by race/ethnicity (Figure 1.1). The highest incidence rates are seen in African Americans (67.0 per 100,000), followed by whites (60.2 per 100,000). African Americans also have a median age at diagnosis that is 4 to 5 years younger than that among whites. Incidence rates among Asian/Pacific Islanders (37.1 per 100,000), American Indians/Alaska Natives (39.9 per 100,000), and Hispanics (30.4 per 100,000) are substantially lower. Globally, lung cancer was responsible for approximately 1.6 million deaths in 2012, making it the leading cause of death due to cancer worldwide (3). Mortality varies with smoking prevalence, with the highest mortality rates in Central and Eastern Europe and Eastern Asia in men and in North America and Northern Europe in women.
SURVIVAL
The 5-year survival rate after a lung cancer diagnosis in the United States was 19.7% for patients diagnosed in 2005 to 2011 (2). While still relatively low, the 5-year survival rate has risen from 12.2% for those diagnosed in 1975 to 1977. Women continue to have higher 5-year survival rates than men (21.2% vs. 15.8%). These poor survival rates reflect the high frequency of late-stage disease at diagnosis, with 57% of lung cancer patients in 2005 to 2011 diagnosed at distant stage with only a 4.2% 5-year survival rate. For the 16% of patients diagnosed with localized disease, the 5-year survival rate was 54.8%. Figure 1.2 illustrates the disparity in survival by race/ethnicity. The 5-year survival rate is highest among Asians (21.2%), followed by whites (20.0%), Hispanics (19.1%), Pacific Islanders (17.5%), African Americans (16.6%), and American Indians/Alaska Natives (14.5%). So, while lung cancer is one of the few cancers for which the major etiologic factor, cigarette smoking, is well known, these statistics reflect the lack of substantial progress that has been made in smoking prevention and cessation, early diagnosis, and treatment.
Figure 1.1 Lung Cancer Incidence and Mortality Rates per 100,000 From 2008–2012 by Race/Ethnicity
Source: From Ref. (2). Howlander N, Krapcho M, Garshell J, et al. (eds). SEER Cancer Statistics Review, 1975-2012. Bethesda, MD: National Cancer Institute; 2015.
RISK FACTORS FOR LUNG CANCER
The risk factors for lung cancer are summarized in Table 1.1.
Tobacco and Inhaled Smoke Exposures
Smoking Prevalence
In 2012, 18.1% of U.S. adults aged 18 years or older were current smokers, with 7.0% of current smokers smoking ≥30 cigarettes per day (4). Mean cigarettes smoked per day was 14.6. Current smoking is more prevalent among men than women (20.5% vs. 15.8%). The prevalence of cigarette smoking varies by race/ethnicity, with rates among whites being 19.7%, African Americans 18.1%, American Indian/Alaska Natives 21.8%, Hispanics 12.5%, and Asians 10.7%. Both current smokers and former smokers continue to be at increased risk of lung cancer.
Figure 1.2 Five-Year Lung Cancer-Specific Survival Rates by Race/Ethnicity
Source: From Ref. (2). Howlander N, Krapcho M, Garshell J, et al (eds). SEER Cancer Statistics Review, 1975-2012. Bethesda, MD: National Cancer Institute; 2015.
Smoking and Lung Cancer Incidence
The association between cigarette smoking and lung cancer is well known, with a landmark report on smoking and health released by the Surgeon General of the U.S. Public Health Service in 1964. At the time of this report, the risk for lung cancer in smoking men was estimated to be approximately 10-fold higher than in never-smoking men, with lower risk among women. Risk was shown to increase with greater amount smoked and decline with cessation of smoking. This report resulted in a steep decline in smoking among men, driving a decline in lung cancer incidence beginning in the mid-1980s. A somewhat slower decline in cigarette use has been reported among women, with incidence rates beginning to decline in the mid-2000s. The incidence rates for lung cancer reflect the change in the use of tobacco, with women now smoking at rates comparable to men and both sexes starting in their teens.
The National Institutes of Health-American Association of Retired Persons (NIH-AARP) Diet and Health Study provides some of the most current data on smoking habits with follow-up for lung cancer in a cohort of 186,057 women and 266,074 men born between 1924 and 1945, who were enrolled in 1995 and followed for 11 years (5). Among women, 17% were current smokers and 44% were never smokers; among men, 13% were current smokers and 26% were never smokers, with most beginning to smoke between ages 15 and 19. Lung cancer incidence rates increased with amount smoked, with rates similar in men and women within comparable cigarette per day categories. In women, smoking 1 to 10, 11 to 20, 21 to 40, and greater than 40 cigarettes per day was associated with relative risks of 12.2, 19.8, 28.4, and 44.2, respectively. Among men in the same cigarette per day categories, relative risks were 17.3, 27.9, 37.3, and 53.0, respectively.
African Americans are more likely than other racial/ethnic groups to smoke menthol cigarettes, with one study reporting that 86% of African Americans smoked menthol cigarettes compared to only 23% of whites (6). Smokers of menthol cigarettes in this study, while at increased risk of lung cancer compared with never smokers, had lower rates of lung cancer than smokers of nonmenthol cigarettes. Other tobacco exposures in the form of cigar and pipe use are associated with a two- to fivefold increase in lung cancer risk (7). This risk is lower than that seen for cigarette smoking, since exposures are typically lower and smoke inhalation is not as deep.
Smoking Cessation
Several studies have shown that quitting smoking is associated with a reduction in lung cancer risk that continues to decrease with years since quitting, although risk never reaches the level among never smokers. Studies in the United Kingdom showed that among men who stopped smoking at ages 60, 50, 40, and 30, the cumulative risks of lung cancer by age 75 were 10%, 6%, 3%, and 2%, respectively, compared to a cumulative risk of death from lung cancer of 16% in men who did not quit (8). Using published data, Fry et al. estimated that after quitting it took 7.9 years in women and 10.7 years in men for the excess lung cancer risk to become half that of a continuing smoker (9).
Environmental Tobacco Smoke
More than two decades after the Surgeon General’s report linking cigarette smoking to lung cancer, a second report from the Surgeon General detailed the risk associated with second-hand smoke, or environmental tobacco smoke (ETS), in nonsmoking individuals. Several known carcinogens in tobacco smoke are also present in ETS including aromatic amines, polycyclic aromatic hydrocarbons, and tobacco-specific nitrosoamines. ETS is associated with a 20% to 40% increase in lung cancer risk (10). While this is a much more moderate risk of lung cancer than that seen with cigarette smoking, ETS still poses a substantial risk.
Tobacco Alternatives
Alternatives to tobacco use include e-cigarettes, marijuana, and hookah use. E-cigarettes deliver aerosolized nicotine. Their use is not yet regulated by the Food and Drug Administration (FDA), so animal and human safety of e-cigarettes has not been evaluated. Little is known about lung cancer risks associated with e-cigarettes given the recent nature of use. What is known is that e-cigarettes deliver similar levels of nicotine as smoking cigarettes, that is enough to evoke physiologic responses in humans. Nicotine and its metabolites are carcinogenic; they contain potentially toxic aldehydes and reactive oxygen species, and alter cytokine levels in murine models.
Risks of lung cancer associated with marijuana use have been difficult to estimate. There is substantial evidence that habitual use is associated with respiratory symptoms, airflow limitation and airway inflammation, and in murine models, marijuana smoke condensate is associated with alterations in gene expression similar to that seen with tobacco smoke condensate (11). In a Swedish study, use of marijuana more than 50 times in men aged 18 to 20 was associated with a twofold increased risk of lung cancer after adjustment for tobacco use (12). Marijuana use is relatively prevalent, with 23.1% of 10- to 24-year olds in the United States in 2011 reporting use within 30 days of being surveyed (13). With new state laws allowing medical and recreational marijuana use, continued study of the long-term effects of this exposure is needed.
Another practice that is increasing in prevalence is hookah use (waterpipe tobacco smoking). In 2011, 18.5% of 12th grade students in the United States had used a hookah in the past year (14), with rates of use in university students ranging from 22% to 40% (15, 16). Exposure to smoke from a hookah results in mean peak plasma nicotine concentrations similar to those resulting from smoking a cigarette, with a 3.75-fold greater increase in carboxyhemoglobin levels and an inhaled smoke volume over 50-fold greater than that of a cigarette (17). The full effects of alternative inhaled smoke will not be fully realized for many decades.
Environmental Exposures
Asbestos
One of the strongest environmental risk factors for lung cancer is asbestos. Asbestos includes several types of naturally occurring mineral fibers that have been widely used in industry. Use peaked in the 1970s, with a subsequent decline in industrial use in the United States due to bans in certain products by the Consumer Product Safety Commission in 1977 and reductions in Occupational Safety and Health Administration (OSHA) guidelines for permissible exposure limits. Asbestos exposure has been linked to increased lung cancer risk in multiple studies, with latency times of 20 to 40 years. Risk is both dose-dependent and related to the size and composition of the inhaled fibers (18). A synergistic effect on lung cancer risk is seen with cigarette smoking (19).
Radon
Radon was first linked to lung cancer among underground miners with high exposures. Outside of occupational settings, individual exposures can occur when radon appears as a contaminant of indoor air. Residential exposures have been associated with an approximately 10% increase in lung cancer risk per 100 Bq/m3 increase in measured radon, with a linear dose–response relationship. Synergistic effects with cigarette smoking result in a 25-fold higher risk among smokers than among non-smokers (20,21). It is estimated that 20,000 lung cancers diagnosed annually in the United States are attributable to radon exposure, prompting the Environmental Protection Agency in 2011 to develop plans for increased awareness and risk reduction (22).
Ionizing Radiation
In studies of Hiroshima and Nagasaki atomic bomb survivors, ionizing radiation has been associated with increased lung cancer risk with a linear dose–response relationship (23). Much lower doses of ionizing radiation, such as those received during diagnostic medical procedures, have not generally been associated with increased lung cancer risk. As lung cancer screening with low-dose CT becomes more prevalent, evaluation of associated lung cancer risk trends should be conducted. The risks associated with repeated medical screening and diagnostic procedures should be weighed against the benefits (24).
Air Pollution
Exposure to indoor air pollution from use of coal for cooking and heating, and the heating of cooking oils has been associated with increased lung cancer risk, particularly in studies conducted in China where such exposures are high. Household coal use was evaluated in a meta-analysis of 25 studies, showing an overall two-fold increased risk of lung cancer with exposure (25). In Chinese nonsmoking women, heating cooking oils to high temperatures is also associated with an increased risk of lung cancer (26). Outdoor air pollution in the form of ambient fine particulate matter (PM2.5) has been linked to lung cancer mortality, but these associations tend to be weak (27,28).
Occupational exposure to diesel fuel is associated with an approximately 30% increase in lung cancer risk, with a significant dose-response trend (29). The risk of lung cancer is increased about 50% among farmers with diesel exposure related to daily tractor use, with higher risk seen for adenocarcinoma (30). While diesel exhaust contains known carcinogens, more study is needed to understand the risks associated with long-term, low-level exposures. Other exposures associated with increased lung cancer risk include wood dust (31) and low-level environmental exposure to cadmium (32).
Genetic Susceptibility
Familial Risk
A hallmark of genetic susceptibility to cancer is aggregation in families. Studying familial aggregation in lung cancer is challenging given that smoking clusters in families and is such a strong risk factor for the disease. Even with these challenges, there is consistent evidence demonstrating a two- to fourfold increased risk associated with having a first-degree relative with lung cancer after accounting for other risk factors, including smoking amount and duration, with variation in risk estimates by age of lung cancer diagnosis, smoking status, and race (33–36). Meta-analyses demonstrate fairly consistent findings, with an approximately 1.5- to 2-fold increased risk of lung cancer associated with family history (37). Variation in risk estimates are seen by race (1.5 for whites, 2.1 for African Americans) and by age at diagnosis (2.0 for onset <50 years).
The level of risk associated with family history is similar to that seen for breast, colon, and prostate cancer, suggesting an underlying genetic contribution to lung cancer susceptibility. Family studies to identify rare, highly penetrant, inherited mutations have been limited to one study that reported the first evidence of a lung cancer susceptibility locus on chromosome 6q23-25 segregating in high-risk lung cancer families (38). As the number of relatives and generations affected with lung cancer increased, so did the significance of this finding. Most importantly, putative carriers of risk in this locus were at higher risk even if they were never smokers or had light smoking histories, suggesting that any level of tobacco exposure increases risk among those with inherited lung cancer susceptibility. A germline mutation in PARK2 in this region was linked to lung cancer risk in one family with eight affected members (39). Additional evidence of linkage was found for regions on chromosomes 1q, 8q, 9p, 12q, 5q, 14q, and 16q (38). A two-stage genome-wide association study (GWAS) that focused on variants associated with a family history of lung cancer, identified SNPs on chromosomes 4p15.2 and 10q23.33 (40). GWAS data have also been used to estimate overall heritability and the proportion of heritability associated with smoking. It is estimated that 24% of the heritability of lung cancer is attributable to genetic determinants of smoking (41).
Genome-Wide Association Studies
Beyond rare, highly penetrant inherited mutations contributing to lung cancer risk, there is also evidence for contributions to susceptibility due to more common, low-penetrant genetic alterations identified by GWAS (Table 1.2). Several early GWAS identified a region on chromosome 15q25 that is more common in lung cancer cases than in controls (42–44). A neuronal nicotinic acetylcholine receptor gene cluster comprising CHRNA3 and CHRNA5 subunits lies within this region. Genetic variation here is associated with an approximate 30% increased risk of lung cancer among individuals carrying a heterozygous mutation and an 80% increase for those who are homozygous for a mutation. A meta-analysis of smokers with or without lung cancer and/or chronic obstructive pulmonary disease (COPD) reported that multiple loci within this region are associated with cigarettes smoked per day and at least one locus is associated with lung cancer independent of amount smoked (45).
Additional regions of interest have been identified on chromosomes 6p21 and 5p15 (42–44). A large meta-analysis of 14,900 lung cancer cases and 29,485 controls of European ancestry from 16 GWAS validated associations between lung cancer risk and genetic variation at 5p15, 6p21, and 15q25 (46). Imputation has been used to expand the data available from GWAS, resulting in the identification and validation of associations between squamous cell carcinoma of the lung and rare variants of BRCA2-K3326X (odds ratio [OR] 2.5, P = 4.7 × 10−20) and CHEK2-I157T (OR 0.38, P = 1.27 × 10−13) (47).
The largest GWAS have included only individuals of European ancestry. The varied genetic architecture and smoking histories of different race/ethnic groups make GWAS in other populations necessary for the eventual identification of lung cancer susceptibility genes. The findings among whites on 15q25, 5p15, and 6p21 have been replicated in African Americans (48, 49). In the Han Chinese population, GWAS identified lung cancer risk associations on 5p15, 3q28, 13q12, and 22q12 (50), and 10p14, 5q32, and 20q13 (51). In the Japanese population, the findings on 5p15, 3q28, and 6p21 were also replicated (52).
GWAS in never smokers has been limited. In never-smoking Asian women, the 6p21, 5p15, and 3q28 findings were replicated and regions on 10q25 and 6q22 were also identified as being associated with lung cancer (53). A large GWAS in never smokers of European ancestry is underway. These studies, across multiple populations and exposure groups, have the potential to uncover mechanisms of carcinogenesis and to identify high-risk individuals to target for prevention and screening efforts.
Chronic Obstructive Pulmonary Disease
COPD and Lung Cancer Incidence
Substantial evidence has been published demonstrating a link between COPD and lung cancer. These diseases share a common risk factor, cigarette smoking, but studies also suggest that a history of COPD is related to a two- to threefold increased risk for lung cancer independent of cigarette smoke exposure (54–57). This link between COPD and lung cancer is evident even among never smokers (58). COPD represents a disease with multiple phenotypes, including emphysema and chronic bronchitis, and it is less clear how lung cancer risk varies by COPD phenotype (57,59). Epidemiologic data relying on self-report of COPD show some differences in risk by COPD phenotype. The largest meta-analysis reported that lung cancer was associated with a previous history of COPD (OR 2.2, 95% CI 1.7–3.0), chronic bronchitis (OR 1.5, 95% CI 1.3–1.8), and emphysema (OR 2.0, 95% CI 1.7–2.4) (56).
Clinical studies have used CT evidence of emphysema and/or spirometry-defined measures of airflow obstruction to evaluate subsequent risk of lung cancer, reducing the potential for disease misclassification and recall bias. Increased risk of lung cancer has been associated with decreasing forced expiratory volume in 1 second (FEV1) even in smokers with only minimal declines in FEV1 (59). Several studies report a two- to fourfold increased risk of lung cancer in the presence of CT evidence of emphysema, with no or lower risk associated with airflow obstruction (55,60–62). In studies using quantitative image analysis of CTs (qCT), CT measures of emphysema have not been associated with lung cancer independent of other COPD measures (63,64). In a meta-analysis including seven studies, Smith et al. (65) reported a threefold (95% CI 2.71, 4.51) increased risk of lung cancer associated with visually detected emphysema, but a nonsignificant 1.16-fold (95% CI 0.48, 2.81) increased risk of lung cancer with qCT-defined emphysema. These studies demonstrate the need to consistently define COPD to better understand the relationship between COPD and lung cancer.
COPD and Genetic Risk
Candidate gene studies in COPD and lung cancer have identified common and shared genetic variation in inflammation, extracellular matrix proteolysis, and oxidative stress pathways (66,67), including SNPs in epoxide hydrolase 1 (EPHX1), matrix metalloproteinases, and interleukin 1β (IL1B) (68–70). Inflammatory pathway gene SNPs in IL7R, IL15, TNF, TNFRSF10A, IL1RN, and IL1A have been associated with lung cancer risk differentially by self-reported history of COPD (71). SNPs in IL1A have also been reported to be more strongly associated with lung cancer risk in those with emphysema (69). Genetic variation on 15q25.1, as reported from GWAS in lung cancer, has been reported in GWAS for COPD-related phenotypes as well (45,72–74). Limited studies have specifically evaluated a joint lung cancer and COPD phenotype. Summary data show that the 15q25 locus is associated with risk of both diseases, genetic variants on 4q31 and 4q22 are associated with reduced risk of both diseases, loci on 6p21 are most strongly associated with lung cancer risk in smokers with COPD, and variants on 5p15 and 1q23 alter lung cancer risk when COPD is not present (75). It is clear that additional work is needed to untangle the COPD-lung cancer relationship.
Infectious Agents
Tuberculosis
Lung cancer risk has been associated with several infectious diseases, including tuberculosis (TB). In a meta-analysis of 37 case-control studies and four cohort studies, TB was associated with a 70% increased risk of lung cancer (95% CI 1.5–2.0), adjusting for smoking. Similar risk is observed among never smokers and the highest risk is seen within 5 years of a TB diagnosis (76).
Human Papillomavirus
Human papillomavirus (HPV) prevalence in lung tumor tissue ranges from 0% to 100%, with differences by geographic regions, histology subtype of lung cancer, sex, and HPV subtype (77,78), but little is known about HPV-lung cancer risk profiles.
Human Immunodeficiency Virus
HIV infection has been associated with a 1.5- to 5.0-fold increased risk of lung cancer. In a review of 65 studies, lung cancer risk in HIV-positive populations varied by geographic region. Standardized incidence ratios or incidence rate ratios were 1.5 to 3.4 in Europe, 0.7 to 6.9 in the United States, and 5.0 in Africa (79). Lung cancer risk among HIV-infected patients receiving highly active antiretroviral therapy (HAART) is similar to that among patients not receiving HAART (80,81). The cumulative incidence of lung cancer by age 75 is 3.4% among those with HIV and 2.8% among those without HIV (82).
SUMMARY
Lung cancer is the leading cause of cancer-related death globally and, while incidence and mortality rates have declined with decreased smoking rates, the 5-year survival rate continues to be less than 20%. Although the major risk factor for this disease is well known (cigarette smoking), progress in prevention and early diagnosis has been slow. In addition to cigarette smoking, lung cancer has been associated with exposure to asbestos, radon, ionizing radiation, and indoor air pollution. A contribution for genetic susceptibility is well described in both family-based studies and GWAS. Continued exploration of the role of COPD, infections and exposure to smoke from marijuana, hookah, and e-cigarettes, is needed to further define lung cancer risk. Well-defined high-risk groups, based on more than just smoking history, will be needed to best target populations for lung cancer prevention and screening.
