Advertisement

The Relationship Between Air Pollution and Lung Cancer in Nonsmokers in Taiwan

Open ArchivePublished:January 18, 2019DOI:https://doi.org/10.1016/j.jtho.2018.12.033

      Abstract

      Introduction

      For never-smokers (smoked <100 lifetime cigarettes), lung cancer (LC) has emerged as an important issue. We aimed to investigate the effects of prevalence changes in tobacco smoking and particulate matter (PM) 2.5 (PM2.5) levels on LC in Taiwan, in relation to contrasting PM2.5 levels, between Northern Taiwan (NT) and Southern Taiwan (ST).

      Methods

      We reviewed 371,084 patients with LC to assess smoking prevalence and correlations between the incidence of adenocarcinoma lung cancer (AdLC) and non-AdLC. Two subsets were selected to assess different AdLC stage trends and the effect of PM2.5 on survival of patients with AdLC.

      Results

      From 1995 to 2015, the proportion of male adult ever-smokers decreased from 59.4% to 29.9% whereas the female smoking rate remained low (3.2% to 5.3%). AdLC incidence in males and females increased from 9.06 to 23.25 and 7.05 to 24.22 per 100,000 population, respectively. Since 1993, atmospheric visibility in NT improved (from 7.6 to 11.5 km), but deteriorated in ST (from 16.3 to 4.2 km). The annual percent change in AdLC stages IB to IV was 0.3% since 2009 (95% confidence interval [CI]: -1.9%–2.6%) in NT, and 4.6% since 2007 (95% CI: 3.3%–5.8%) in ST; 53% patients with LC had never smoked. Five-year survival rates for never-smokers, those with EGFR wild-type genes, and female patients with AdLC were 12.6% in NT and 4.5% in ST (hazard ratio: 0.79, 95% CI: 0.70–0.90).

      Conclusions

      In Taiwan, greater than 50% of patients with LC had never smoked. PM2.5 level changes can affect AdLC incidence and patient survival.

      Keywords

      Introduction

      In the United states and in other countries where tobacco smoking has been common it was previously thought that 90% of lung cancer (LC) resulted solely from cigarette smoking.
      • Alberg A.J.
      • Ford J.G.
      • Samet J.M.
      American College of Chest Physicians. Epidemiology of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition).
      However, in 2016, Whiteman and Wilson
      • Ferlay J.
      • Soerjomataram I.
      • Dikshit R.
      • et al.
      Cancer incidence and mortality worldwide: sources, methods, and major patterns in GLOBOCAN 2012.
      revealed that the global tobacco population attributable fraction (PAF) was high in men (81%) but markedly lower in women (58%).
      • Whiteman D.C.
      • Wilson L.F.
      The fractions of cancer attributable to modifiable factors: a global review.
      According to the GLOBOCAN 2012 study, the relatively high rate of LC in east Asian women is of particular interest because tobacco smoking is generally rare in these populations.
      • Ferlay J.
      • Soerjomataram I.
      • Dikshit R.
      • et al.
      Cancer incidence and mortality worldwide: sources, methods, and major patterns in GLOBOCAN 2012.
      Compared to LC rates among smokers, LC in those who have never smoked is associated with patients of east Asian descent, females, and those with an adenocarcinoma lung cancer (AdLC) histology.
      • Thu K.L.
      • Vucic E.A.
      • Chari R.
      • et al.
      Lung adenocarcinoma of never smokers and smokers harbor differential regions of genetic alteration and exhibit different levels of genomic instability.
      • Toh C.K.
      • Gao F.
      • Lim W.T.
      • et al.
      Never-smokers with lung cancer: epidemiologic evidence of a distinct disease entity.
      This population is more likely to have distinct molecular markers, especially EGFR, ALK receptor tyrosine kinase (ALK), and ROS1 gene mutations. Comprehensive genomic analysis using whole-genome sequencing has identified significant differences between the tumor genome of LC in patients who have never smoked and smokers.
      • Govindan R.
      • Ding L.
      • Griffith M.
      • et al.
      Genomic landscape of non–small cell lung cancer in smokers and never-smokers.
      Particulate matter (PM) air pollution contributes to the incidence of AdLC in Europe, but the corresponding effect in East Asia is uncertain.
      • Raaschou-Nielsen O.
      • Andersen Z.J.
      • Beelen R.
      • et al.
      Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE).
      In Taiwan, there have been differences in air pollution trends between Northern (NT) and Southern (ST) Taiwan for approximately 25 years. Using both the Taiwan National Lung Cancer database Adult Smoking Behavior Surveillance System and the long-term nationwide visibility data, we aimed to investigate the effect of air pollution level changes on LC rates in Taiwan, where there is a high prevalence rate of LC among patients who have never smoked.

      Methods

      Since 1970, the Taiwan Central Weather Bureau has monitored long-term nationwide atmospheric visibility data. To determine a more precise association between visibility and particulate matter less than 2.5 μm (PM2.5), we selected daytime visibility data to represent daily PM2.5 including only day-time humidity levels lower than 70%, and excluding the daily rainfall accumulation higher than 0 cm.
      • Abbey D.E.
      • Ostro B.E.
      • Fraser G.
      • Vancuren T.
      • Burchette R.J.
      Estimating fine particulates less than 2.5 microns in aerodynamic diameter (PM2.5) from airport visibility data in California.
      • Guo L.L.
      • Zheng H.
      • Lyu Y.L.
      • Liu L.Y.
      • Kong F.
      • Wang S.R.
      Trends in atmospheric particles and their light extinction performance between 1980 and 2015 in Beijing, China.
      Since 2004, the Environmental Protection Administration has monitored PM2.5 levels across all of Taiwan, and we determined the relationship between atmospheric visibility and PM2.5 (using monitoring data) from 2004 to 2010 as follows: PM2.5=430.523.2+visibility(kilometer)(Supplemental Fig. 1).
      • Guo L.L.
      • Zheng H.
      • Lyu Y.L.
      • Liu L.Y.
      • Kong F.
      • Wang S.R.
      Trends in atmospheric particles and their light extinction performance between 1980 and 2015 in Beijing, China.
      Our study cohorts and the two subsets are described below.

      Study Cohort (1995–2015) and the Adult Smoking Behavior Surveillance System

      This was a retrospective cohort study using the national Taiwan Cancer Registry (TCR) data, where all cancer cases in Taiwan have been recorded in a uniform format since 1979 (Supplemental Fig. 2). In 1995, national health insurance was instituted and TCR data were closed to public use. To determine cancer trends and any correlation with smoking prevalence, the histology results of a cohort from 1995 to 2015 were used. Clinical data for analysis included patient age, sex, and histological type. Age-standardized rates of incidence were calculated based on the direct method, using the 2000 world standard population as defined by the WHO and expressed as cases per 100,000 population.
      Department of Household Registration, Ministry of the Interior
      Population by sex and 5-year age group.
      The Taiwan Adult Smoking Behavior Survey, initiated in 1990, was originally sponsored by the Taiwan Tobacco & Liquor Corporation.
      Health Promotion Administration, Ministry of Health and Welfare, Taiwan
      Adult Smoking Behavior Survey.
      As part of this survey, household interviews were conducted and smoking behavior data were obtained through a questionnaire. In 2004, the Taiwan Health Promotion Administration established the Adult Smoking Behavior Surveillance System. On average, 20,000 adults were randomly sampled according to the probability proportional to size method using telephone numbers with an even distribution across 25 counties. The interview questionnaire was disseminated using a computer-assisted telephone interviewing system, and each code was verified manually. We compared the Adult Smoking Behavior Survey and cancer trends, categorized according to histology results, between 1990 and 2015.

      Cohort Subset I (2004–2015) and Visibility Trends in NT and ST (1980–2004)

      After the Cancer Control Act was enacted in 2003, the national cancer registration rate was greater than 92%, and cancer staging and complete data were more reliable.
      • Chiang C.J.
      • Chen Y.C.
      • Chen C.J.
      • et al.
      Cancer trends in Taiwan.
      • Chiang C.J.
      • Lo W.C.
      • Yang Y.W.
      • You S.L.
      • Chen C.J.
      • Lai M.S.
      Incidence and survival of adult cancer patients in Taiwan, 2002–2012.
      Cohort subset I (2004–2015) was used to determine LC trends according to different stage groups and any correlation with visibility. Recent LC incidence rates might be affected due to low-dose computed tomography (LDCT) screening in stage IA patients with LC. In cohort subset I, we selected only stages IB to IV or stages IIIB to IV AdLC cases for sensitivity analysis. Trends in the annual age-standardized rates were analyzed according to the annual percent change (APC) using joinpoint regression analysis. The visibility trend in the 25-year period before a diagnosis of cancer in cohort subset I was used as a surrogate for the air pollution trend.
      • Abbey D.E.
      • Ostro B.E.
      • Fraser G.
      • Vancuren T.
      • Burchette R.J.
      Estimating fine particulates less than 2.5 microns in aerodynamic diameter (PM2.5) from airport visibility data in California.
      • Guo L.L.
      • Zheng H.
      • Lyu Y.L.
      • Liu L.Y.
      • Kong F.
      • Wang S.R.
      Trends in atmospheric particles and their light extinction performance between 1980 and 2015 in Beijing, China.
      Six stations in Taipei and two stations in Kaohsiung monitoring air quality zones were used to determine the visibility for NT and ST (Supplemental Fig. 3). The postcode areas for cancer incidence in NT included 100–199, 200–208, 220–259, and 320–338, and those for ST included 800–859 and 900–948.

      Cohort Subset II (2011–2015), the PAF, and Overall Survival

      To further evaluate the effect of smoking on LC and the visibility effect on cancer survival, we required a detailed smoking history and the EGFR mutation status of patients with LC. Since 2011, the detailed smoking history and EGFR mutation status of patients with LC have been routinely recorded in the TCR. Never-smokers were defined as those who had smoked less than 100 cigarettes in their lifetime, whereas others were defined as those who had ever smoked (ever smokers). Cohort subset II (2011–2015) was used for the PAF calculation. The PAF was used to identify the number of cancer cases that might have been avoided if underlying causal factors had not occurred. PAFs were calculated using the following formula:PAF=p×(IRR1)p×(IRR1)+1×100, where p is the smoking prevalence in the total population (>20 years old), and IRR is the incidence rate ratio of LC for smoking.
      Overall survival (OS) was estimated using the Kaplan-Meier method, whereas between-group differences in OS were assessed using the stratified log-rank test. Hazard ratios (HRs) and associated 95% confidence intervals (CIs) were estimated using the Cox proportional hazards model. All analyses were performed using SAS version 9.4 statistical software (SAS Institute, Cary, North Carolina). This study was approved by the Institutional Review Board of Taichung Veterans General Hospital (IRB No.: CE17068B).

      Results

      The Entire Cohort

      In total, 371,084 patients with LC were registered from 1995 to 2015. Figure 1 shows the combined proportion of adult ever-smokers in Taiwan and LC trends according to histology results for the 25-year period. The proportion of male adult ever-smokers in Taiwan decreased from 59.4% to 29.9% (Fig. 1A), and the age-adjusted incidence of non-AdLC peaked in 2003 and decreased thereafter. However, the rate of AdLC among males continued to increase, growing from 9.06 to 23.25 per 100,000 population from 1995 to 2015. Among adult females, the proportion of ever-smokers remained consistently low at approximately 3.2% to 5.3%, and non-AdLC remained less than 15 per 100,000 population from 1995 to 2015 (Fig. 1B). However, the age-adjusted incidence rate of AdLC in females steadily increased from 7.05 to 24.22 per 100,000 population during the same period.
      Figure thumbnail gr1
      Figure 1Trends in smoking percentage in the adult population (A: Male; B: Female) from 1990 to 2015 (dotted line), and estimated age-adjusted incidence rates for adenocarcinoma lung cancer (dark solid line) and non-adenocarcinoma lung cancer (light solid line) from 1995 to 2015.

      Cohort Subset I

      Before 1992, Taipei city and NT had poor atmospheric visibility (ranging from 6.5 km to 8.0 km) because of the growing population, traffic volumes, and industrial factories. Visibility in Taipei improved between 1993 and 2003 to 11.5 km because of the construction and expansion of a mass transit rail system in Taipei, and because factories were moved outside Taipei (Fig. 2A, Supplemental Fig. 3).
      • Tsai Y.I.
      • Kuo S.C.
      • Lee W.J.
      • Chen C.L.
      • Chen P.T.
      Long-term visibility trends in one highly urbanized, one highly industrialized, and two rural areas of Taiwan.
      In contrast, since 1970, the petrochemical industry has been rapidly developing in ST, and atmospheric visibility declined between 1980 and 2000 (from 20.1 km to 4.2 km) (Fig. 2A, Supplemental Fig. 3). Annual events days of atmospheric visibility less than 10 km or 5 km showed consistent trends (Fig. 2B, Supplemental Table 1). The incidence rates for AdLCs, including stages IB to IV and stages IIIB to IV in NT and ST, are shown in Figures 2C and D, respectively. We selected the best model according to the results from joinpoint regression analysis. In Figure 2C, the APC in stages IB to IV AdLC was 0.3% (95% CI: -1.9%–2.6%) from 2009 in NT, and 4.6% (95% CI: 3.3%–5.8%) from 2007 in ST. In Figure 2D, we selected only advanced AdLC (stages IIIB and IV), and the APC was 0.1% (95% CI: -2.2% to 2.1%) from 2008 in NT and 3.8% (95% CI: 2.3%–5.3%) from 2007 in ST.
      Figure thumbnail gr2
      Figure 2Trends in atmospheric visibility from 1980 to 2005 and age-adjusted incidence of lung adenocarcinoma from 2004 to 2015. (A) Visibility (km); (B) proportion of days with a visibility less than 10 km (%); (C) age-adjusted incidence of stages IB to IV adenocarcinoma lung cancer; and (D) age-adjusted incidence of stages IIIB to IV adenocarcinoma lung cancer in northern Taiwan and southern Taiwan.

      Cohort Subset II

      Of 51,933 patients with newly diagnosed LC in Taiwan from 2011 to 2015, 51,416 (99.0%) with recorded smoking behavior data were enrolled for PAF analysis. Of these, 27,236 (53.0%) had never smoked, with 25.5% and 92.1% of male and female patients, respectively, having never smoked; 24,180 (47.0%) were ever-smokers, among whom 6,287 (12.2%) were current smokers with more than 30 pack-years, and 3,670 (7.1%) were former smokers with more than 30 pack-years, and had quit smoking less than 15 years ago (Fig. 3). The patients were divided into never-smoker and ever-smoker (current plus previous smoker) groups, and patient characteristics are shown in Supplemental Table 2. Stage IV AdLC was observed in 58.7% of never-smokers and in 59.3% of ever-smokers. In the never-smoker group, AdLC, squamous cell carcinoma, and small cell carcinoma accounted for 83.5%, 6.2%, and 2.1%, respectively; in ever-smokers, the corresponding values were 48.7%, 25.8%, and 13.6%, respectively. The EGFR-positive mutation rate in stages IIIB and IV AdLC patients was 60.0% in never-smokers and 35.6% in ever-smokers. The clinical characteristics according to histology results in patients with LC are listed in Supplemental Table 3.
      Figure thumbnail gr3
      Figure 3Smoking behavior in Taiwan lung cancer patients from 2011 to 2015 (n = 51,416)
      The standardized rate ratios for male ever-smokers compared to male never-smokers were 1.7, 5.4, and 6.9 in adenocarcinoma, squamous cell carcinoma, and small cell carcinoma, respectively; the corresponding female ever-smoker values compared to female never-smokers values were 2.2, 10.9, and 43.1, respectively (Table 1). Considering the lag effect of cigarette smoking on LC development, the proportion of the population that ever smoked was defined in three different ways: (1) the proportion of current smokers in the population during the same period (2011–2015); (2) current and former smokers during the same period (2011–2015), and; (3) actively having smoked for 20 years before the period 1990–1995. The PAF of tobacco smoking was between 59.9% and 76.7% in males with non-AdLC, between 17.9% and 26.8% in males with AdLC, between 29.9% and 50.1% in females with non-AdLC, and between 3.8% and 4.8% in females with AdLC.
      Table 1Estimated Age-Adjusted Incidence Rates and Standardized Rate Ratios of Ever Smokers Versus Never Smokers, and Population Attributable Fractions for Lung Cancer, by Sex and Histologic Type
      SexHistologic TypeEAIR (SE)SRR = (EAIR in ever smokers) / (EAIR in never smokers)Population Attributable Fraction of Smoking on EAIR (%)
      Ever SmokersNever Smokers2011–2015
      The proportions of adult current smokers from 2011 to 2015 were 33.7% (male) and 4.3% (female).
      2011–2015
      The proportions of adult current smokers and those who were former smokers from 2011 to 2015 were 51.4% (male) and 6.3% (female).
      1990–1995
      The proportions of adult current smokers from 1990 to 1995 were 56.5% (male) and 3.4% (female).
      MaleAll lung cancer80.7 (0.6)32.3 (0.4)2.5 (2.4–2.6)33.643.645.9
      Adenocarcinoma37.6 (0.4)22.8 (0.3)1.7 (1.6–1.7)17.925.026.8
      Squamous-cell carcinoma22.2 (0.3)4.1 (0.1)5.4 (5.1–5.8)59.969.471.4
      Small-cell carcinoma11.1 (0.2)1.6 (0.1)6.9 (6.2–7.6)66.475.076.7
      FemaleAll lung cancer110.0 (2.2)35.3 (0.3)3.1 (2.9–3.3)8.411.86.8
      Adenocarcinoma66.9 (1.7)31.0 (0.3)2.2 (2.0–2.3)4.86.83.8
      Squamous-cell carcinoma14.8 (0.8)1.4 (0.1)10.9 (8.2–14.4)29.938.525.3
      Small-cell carcinoma15.5 (0.8)0.4 (0.1)43.1 (28.9–64.2)64.572.759.1
      EAIR, estimated age-adjusted incidence rate per 100,000 population ± standard error (SE); SRR, standardized rate ratios.
      a The proportions of adult current smokers from 2011 to 2015 were 33.7% (male) and 4.3% (female).
      b The proportions of adult current smokers and those who were former smokers from 2011 to 2015 were 51.4% (male) and 6.3% (female).
      c The proportions of adult current smokers from 1990 to 1995 were 56.5% (male) and 3.4% (female).
      For 15,972 patients with newly diagnosed stages IB to IV AdLC in NT and ST from 2011 to 2015, the detailed characteristics are listed in Supplemental Table 4. LC among females who had never smoked was higher in ST (93.8%) than in NT (90.2%), and the four groups were similar in terms of age, Eastern Cooperative Oncology Group performance status, smoking behavior, and EGFR mutation status. In Taiwan, smoking behavior, sex, and EGFR mutation status were highly correlated; therefore, instead of using regression analysis, we analyzed OS in stages IIIB to IV AdLC between NT and ST in a subgroup analysis (Table 2). In the OS subgroup analysis, only EGFR wild-type, female sex, and never-smoked status significantly differed between the two areas. The 5-year survival rates in the never smoked, EGFR wild-type, and female subgroups were 12.6% in NT and 4.5% in ST (HR: 0.79, 95% CI: 0.70–0.90).
      Table 2Survival in Stages IIIB to IV Lung Adenocarcinoma Patients From Northern and Southern Taiwan
      Sex/Smoking StatusEGFR

      Mutation
      LocationNo. of Cases1-Year

      Survival
      2-Year

      Survival
      5-Year

      Survival
      LR TestHazard

      Ratio
      Female/never smokedPositiveNorthern Taiwan241174.047.715.40.3710.96 (0.88–1.05)
      Southern Taiwan121574.146.813.9
      NegativeNorthern Taiwan86555.534.612.60 <.0010.79 (0.70–0.90)
      Southern Taiwan46347.126.74.5
      Female/ever smokersPositiveNorthern Taiwan18365.641.119.10.6620.92 (0.64–1.35)
      Southern Taiwan5373.631.715.4
      NegativeNorthern Taiwan15441.618.55.60.8620.96 (0.66–1.47)
      Southern Taiwan3336.420.0NA
      Male/never smokedPositiveNorthern Taiwan71068.639.910.70.9971.00 (0.86–1.16)
      Southern Taiwan35768.640.99.5
      NegativeNorthern Taiwan35645.223.17.40.5911.06 (0.87–1.29)
      Southern Taiwan17346.828.14.3
      Male/ever smokersPositiveNorthern Taiwan99766.136.08.00.0630.89 (0.78–1.01)
      Southern Taiwan46361.130.78.7
      NegativeNorthern Taiwan135036.416.85.30.1930.94 (0.85–1.04)
      Southern Taiwan60736.213.94.6
      LR, likelihood ratio; NA, not available because of small sample size in this group.

      Discussion

      This nationwide study linked the risk factors for LC and showed that patients who never smoked comprised a major proportion of patients with LC in Taiwan. This latter factor appears to have highly affected the rapid increase in the AdLC incidence, especially in ST. The standardized rate ratio and PAF for smoking were both low in with AdLC. Further, approximately 60% of ever-smoker and never-smoker patients with AdLC already had stage IV disease.
      According to the National Lung Screening Trial criteria and an expert panel report, annual LDCT screening should be offered only to current or former smokers who had quit smoking within the previous 15 years and who had 30 pack-years or more of cigarette smoking.
      • National Lung Screening Trial Research T.
      • Aberle D.R.
      • Adams A.M.
      • et al.
      Reduced lung-cancer mortality with low-dose computed tomographic screening.
      • Mazzone P.J.
      • Silvestri G.A.
      • Patel S.
      • et al.
      Screening for lung cancer: CHEST guideline and expert panel report.
      Based on this recommendation, less than 20% of our patients with LC would have been included in the screening; this is common throughout East Asia.
      • Wakai K.
      • Ando M.
      • Ozasa K.
      • et al.
      Updated information on risk factors for lung cancer: findings from the JACC Study.
      Other high-risk factors should be selected for early LC screening including blood biomarkers such as plasma microRNA signatures, autoantibodies to LC-associated antigens, and plasma-protein assay using mass spectrometry.
      • Kosaka T.
      • Yatabe Y.
      • Endoh H.
      • Kuwano H.
      • Takahashi T.
      • Mitsudomi T.
      Mutations of the epidermal growth factor receptor gene in lung cancer: biological and clinical implications.
      IARC Working Group on the Evaluation of Carcinogenic Risks to Humans
      Tobacco smoke and involuntary smoking.
      • Couraud S.
      • Debieuvre D.
      • Moreau L.
      • et al.
      No impact of passive smoke on the somatic profile of lung cancers in never-smokers.
      Among Japanese patients with LC, 83% of female patients and 10% of male patients had never smoked.
      • Kosaka T.
      • Yatabe Y.
      • Endoh H.
      • Kuwano H.
      • Takahashi T.
      • Mitsudomi T.
      Mutations of the epidermal growth factor receptor gene in lung cancer: biological and clinical implications.
      In Taiwan, these values were higher (92.1% of female and 25.5% of male patients had never smoked). Additionally, in Taiwan, the smoking rate among females has remained at approximately 4% during the past 25 years, and the smoking rate among males had decreased from approximately 60% to 30% within the same period; however, the AdLC incidence rate is still increasing in both sexes. Environmental tobacco smoke is a relatively weak carcinogen and accounts for only a minority of LC in patients who have never smoked. Although passive smoking has also been implicated in LC among patients who have never smoked, Couraud et al.
      • Couraud S.
      • Debieuvre D.
      • Moreau L.
      • et al.
      No impact of passive smoke on the somatic profile of lung cancers in never-smokers.
      did not show any clear association between passive smoke exposure and a somatic profile in LC among patients who had never smoked.
      IARC Working Group on the Evaluation of Carcinogenic Risks to Humans
      Tobacco smoke and involuntary smoking.
      • Couraud S.
      • Debieuvre D.
      • Moreau L.
      • et al.
      No impact of passive smoke on the somatic profile of lung cancers in never-smokers.
      In East Asia, there are several unique characteristics pertaining to LC, such as the predominance of an AdLC histology, the presence of a large proportion of patients who had never smoked, female sex, and the presence of EGFR mutations.
      • Sun S.
      • Schiller J.H.
      • Gazdar A.F.
      Lung cancer in never smokers — a different disease.
      • Bell D.W.
      • Brannigan B.W.
      • Matsuo K.
      • et al.
      Increased prevalence of EGFR-mutant lung cancer in women and in East Asian populations: analysis of estrogen-related polymorphisms.
      However, there is a need for more clarity on the carcinogenesis of EGFR mutations, although it has been identified as a predictive marker for EGFR-ROS1 inhibitor treatment.
      The possible causes of LC among patients who have never smoked, aside from secondhand smoke, include factors such as exposure to radon, asbestos, air pollution, diesel exhaust, cooking oil, and occupational carcinogens, as well as inherited susceptibility
      • Sun S.
      • Schiller J.H.
      • Gazdar A.F.
      Lung cancer in never smokers — a different disease.
      • Alberg A.J.
      • Brock M.V.
      • Ford J.G.
      • Samet J.M.
      • Spivack S.D.
      Epidemiology of lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines.
      ; however, the precise cause for individual cases is often uncertain. Long-term exposure to fine particulate air pollution has genotoxic and mutagenic effects, as shown in laboratory studies, and may lead to increased LC risk through inflammatory injury, reactive oxygen species production, and oxidative damage to DNA.
      • Valavanidis A.
      • Fiotakis K.
      • Vlachogianni T.
      Airborne particulate matter and human health: toxicological assessment and importance of size and composition of particles for oxidative damage and carcinogenic mechanisms.
      • Andre V.
      • Billet S.
      • Pottier D.
      • et al.
      Mutagenicity and genotoxicity of PM2.5 issued from an urbano-industrialized area of Dunkerque (France).
      • Moller P.
      • Jacobsen N.R.
      • Folkmann J.K.
      • et al.
      Role of oxidative damage in toxicity of particulates.
      An evaluation by the International Agency for Research on Cancer showed an increase in the risk of LC with increasing exposure levels to PM and air pollution.
      Air Pollution and Cancer, IARC Scientific Publication No. 161, P8-P9.
      According to a prospective analysis by the European Study of Cohorts for Air Pollution Effects (ESCAPE), for every increase of 5 μg/m3 in the PM2.5 level in the environment, the risk of LC increases by 18%; this is associated with an HR for AdLC of 1.55 (95% CI: 1.05–2.29).
      • Raaschou-Nielsen O.
      • Andersen Z.J.
      • Beelen R.
      • et al.
      Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE).
      Although the risk of LC associated with air pollution is much lower than that associated with smoking (relative risk, 23.3 for male current smokers and 12.7 for female current smokers), everyone is exposed to air pollution.
      Since 2008, the Environmental Protection Administration has been monitoring PM2.5 levels on the entire island of Taiwan, and the average PM2.5 levels in NT between 2008 and 2011 were approximately 27.5 to 25.6 μg/m3 (95% CI), whereas these levels were higher at 38.2 to 42.8 μg/m3 in ST (Supplemental Table 5). These are much higher than the levels to which most of the ESCAPE cohort (6.6 to 31.9 μg/m3) had been exposed.
      • Raaschou-Nielsen O.
      • Andersen Z.J.
      • Beelen R.
      • et al.
      Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE).
      The average PM2.5 levels from 2008 to 2017 were consistent with visibility trends from 1980 to 2004 in NT and ST. We were able to calculate the PM2.5 ranges from 1980 to 2008 to be approximately 29.46 to 44.36 μg/m3 in NT and 21.27 to 58.29 μg/m3 in ST. The visibility trends differed between NT and ST, and the trend of age-adjusted AdLC incidence was stable in NT, but continued to increase in ST, especially with regard to stages IB to IV or stages IIIB to IV AdLC.
      Thus, the public health effect is large in Taiwan and in other countries with a similarly high PM2.5 level. The WHO estimated that, in 2004, smoking caused 5.1 million deaths and 71% of all LC worldwide in 2004, whereas air pollution caused 1.2 million deaths and 8% of LC worldwide.
      World Health Organization
      Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks.
      Therefore, air pollution may need to be added to the list of causes of LC. Furthermore, it should be recognized that air pollution has a large effect on public health. This could be the reason for the high incidence of LC among the population that has never smoked in Taiwan. Additionally, the LDCT screening policy should be modified due to regional differences.
      In this study, among stages IIIB and IV AdLC cases, patients with the EGFR wild-type, female sex, and those with a never-smoked status had 5-year survival rates of 12.6% in NT and 4.5% in ST (HR: 0.79). Subgroup analysis revealed that the survival difference was more prominent in women; however, it was not prominent in EGFR-mutant AdLC because the benefit of treatment with EGFR–tyrosine kinase inhibitors is greater than the risk of air pollution on survival in patients with late-stage LC. Similarly, Turner et al.
      • Turner M.C.
      • Krewski D.
      • Pope 3rd, C.A.
      • Chen Y.
      • Gapstur S.M.
      • Thun M.J.
      Long-term ambient fine particulate matter air pollution and lung cancer in a large cohort of never-smokers.
      examined the association between the mean long-term ambient PM2.5 concentrations and LC-related mortality in a 26-year prospective study of a large cohort of patients who had never smoked. Each 10 μg/m3 increase in the PM2.5 concentration was associated with a 15% to 27% increase in LC-related mortality.
      • Turner M.C.
      • Krewski D.
      • Pope 3rd, C.A.
      • Chen Y.
      • Gapstur S.M.
      • Thun M.J.
      Long-term ambient fine particulate matter air pollution and lung cancer in a large cohort of never-smokers.
      These findings suggest that ambient concentrations of PM2.5 are associated with increases in LC-related mortality.
      The strength of our study is that we used a nationwide database covering almost all cases of cancer registered since the database inception in 1990. We also collected detailed data on cancer staging from 2004, as well as smoking status and EGFR mutation status from 2011. The smoking registration rate was 99%, and EGFR mutation status was recorded in almost 80% of all patients with AdLC.
      • Tseng C.H.
      • Chiang C.J.
      • Tseng J.S.
      • et al.
      EGFR mutation, smoking, and gender in advanced lung adenocarcinoma.
      Our study limitations included the retrospective nature of the study, and possible inaccurate staging and mutation reporting in the database. Moreover, LC prevalence may increase as a consequence of LDCT screening, especially in stage IA patients; however, we only selected stages IB to IV and IIIB to IV AdLC patients, as shown in Figures 2C and D, and the trend was consistent. In this study, we used atmospheric visibility as a surrogate for air pollution, and previous studies have shown that the correlation between them is reliable.
      • Abbey D.E.
      • Ostro B.E.
      • Fraser G.
      • Vancuren T.
      • Burchette R.J.
      Estimating fine particulates less than 2.5 microns in aerodynamic diameter (PM2.5) from airport visibility data in California.
      • Liu M.
      • Bi J.
      • Ma Z.
      Visibility-based PM2.5 concentrations in China: 1957–1964 and 1973–2014.
      Other possible carcinogens, such as radon, or asbestos, were not included in this study.
      In conclusion, patients who had never smoked comprised a major proportion of the patients with LC in Taiwan, and AdLC was the predominant subtype. Nearly 60% of patients with LC who had never smoked had stage IV disease, which was similar to that observed among smokers with LC. Air pollution level changes could affect the incidence of AdLC and possibly influence survival rates for female patients with EGFR wild-type AdLC who have never smoked.
      As early detection is challenging for patients with LC who have never smoked, different strategies should be considered in LDCT screening. Regional differences may need to be considered, especially the PM level. Future LC prevention strategies should also focus on a national policy for air pollution reduction, the manufacture of devices to prevent exposure to PM2.5, and specific measures at the individual level to reduce exposure.

      Acknowledgments

      The study was supported by a grant from the Health and Welfare Surcharge of Tobacco products provided by the Health Promotion Administration, Ministry of Health and Welfare, Taiwan (R.O.C.).

      Supplementary Data

      References

        • Alberg A.J.
        • Ford J.G.
        • Samet J.M.
        American College of Chest Physicians. Epidemiology of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition).
        Chest. 2007; 132: 29S-55S
        • Ferlay J.
        • Soerjomataram I.
        • Dikshit R.
        • et al.
        Cancer incidence and mortality worldwide: sources, methods, and major patterns in GLOBOCAN 2012.
        Int J Cancer. 2015; 136: E359-E386
        • Whiteman D.C.
        • Wilson L.F.
        The fractions of cancer attributable to modifiable factors: a global review.
        Cancer Epidemiol. 2016; 44: 203-221
        • Thu K.L.
        • Vucic E.A.
        • Chari R.
        • et al.
        Lung adenocarcinoma of never smokers and smokers harbor differential regions of genetic alteration and exhibit different levels of genomic instability.
        PLoS One. 2012; 7: e33003
        • Toh C.K.
        • Gao F.
        • Lim W.T.
        • et al.
        Never-smokers with lung cancer: epidemiologic evidence of a distinct disease entity.
        J Clin Oncol. 2006; 24: 2245-2251
        • Govindan R.
        • Ding L.
        • Griffith M.
        • et al.
        Genomic landscape of non–small cell lung cancer in smokers and never-smokers.
        Cell. 2012; 150: 1121-1134
        • Raaschou-Nielsen O.
        • Andersen Z.J.
        • Beelen R.
        • et al.
        Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE).
        Lancet Oncol. 2013; 14: 813-822
        • Abbey D.E.
        • Ostro B.E.
        • Fraser G.
        • Vancuren T.
        • Burchette R.J.
        Estimating fine particulates less than 2.5 microns in aerodynamic diameter (PM2.5) from airport visibility data in California.
        J Expo Anal Environ Epidemiol. 1995; 5: 161-180
        • Guo L.L.
        • Zheng H.
        • Lyu Y.L.
        • Liu L.Y.
        • Kong F.
        • Wang S.R.
        Trends in atmospheric particles and their light extinction performance between 1980 and 2015 in Beijing, China.
        Chemosphere. 2018; 205: 52-61
        • Department of Household Registration, Ministry of the Interior
        Population by sex and 5-year age group.
        https://www.prb.org/2010wpds/
        Date: 2018
        Date accessed: September 20, 2018
        • Health Promotion Administration, Ministry of Health and Welfare, Taiwan
        Adult Smoking Behavior Survey.
        https://www.hpa.gov.tw/EngPages/Detail.aspx?nodeid=1077&pid=6202
        Date: 2018
        Date accessed: September 20, 2018
        • Chiang C.J.
        • Chen Y.C.
        • Chen C.J.
        • et al.
        Cancer trends in Taiwan.
        Jpn J Clin Oncol. 2010; 40: 897-904
        • Chiang C.J.
        • Lo W.C.
        • Yang Y.W.
        • You S.L.
        • Chen C.J.
        • Lai M.S.
        Incidence and survival of adult cancer patients in Taiwan, 2002–2012.
        J Formos Med Assoc. 2016; 115: 1076-1088
        • Tsai Y.I.
        • Kuo S.C.
        • Lee W.J.
        • Chen C.L.
        • Chen P.T.
        Long-term visibility trends in one highly urbanized, one highly industrialized, and two rural areas of Taiwan.
        Sci Total Environ. 2007; 382: 324-341
        • National Lung Screening Trial Research T.
        • Aberle D.R.
        • Adams A.M.
        • et al.
        Reduced lung-cancer mortality with low-dose computed tomographic screening.
        N Engl J Med. 2011; 365: 395-409
        • Mazzone P.J.
        • Silvestri G.A.
        • Patel S.
        • et al.
        Screening for lung cancer: CHEST guideline and expert panel report.
        Chest. 2018; 153: 954-985
        • Wakai K.
        • Ando M.
        • Ozasa K.
        • et al.
        Updated information on risk factors for lung cancer: findings from the JACC Study.
        J Epidemiol. 2005; 15: S134-S139
        • Kosaka T.
        • Yatabe Y.
        • Endoh H.
        • Kuwano H.
        • Takahashi T.
        • Mitsudomi T.
        Mutations of the epidermal growth factor receptor gene in lung cancer: biological and clinical implications.
        Cancer Res. 2004; 64: 8919-8923
        • IARC Working Group on the Evaluation of Carcinogenic Risks to Humans
        Tobacco smoke and involuntary smoking.
        IARC Monogr Eval Carcinog Risks Hum. 2004; 83: 1-1438
        • Couraud S.
        • Debieuvre D.
        • Moreau L.
        • et al.
        No impact of passive smoke on the somatic profile of lung cancers in never-smokers.
        Eur Respir J. 2015; 45: 1415-1425
        • Sun S.
        • Schiller J.H.
        • Gazdar A.F.
        Lung cancer in never smokers — a different disease.
        Nat Rev Cancer. 2007; 7: 778-790
        • Bell D.W.
        • Brannigan B.W.
        • Matsuo K.
        • et al.
        Increased prevalence of EGFR-mutant lung cancer in women and in East Asian populations: analysis of estrogen-related polymorphisms.
        Clin Cancer Res. 2008; 14: 4079-4084
        • Alberg A.J.
        • Brock M.V.
        • Ford J.G.
        • Samet J.M.
        • Spivack S.D.
        Epidemiology of lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines.
        Chest. 2013; 143: e1S-e29
        • Valavanidis A.
        • Fiotakis K.
        • Vlachogianni T.
        Airborne particulate matter and human health: toxicological assessment and importance of size and composition of particles for oxidative damage and carcinogenic mechanisms.
        J Environ Sci Health C Environ Carcinog Ecotoxicol Rev. 2008; 26: 339-362
        • Andre V.
        • Billet S.
        • Pottier D.
        • et al.
        Mutagenicity and genotoxicity of PM2.5 issued from an urbano-industrialized area of Dunkerque (France).
        J Appl Toxicol. 2011; 31: 131-138
        • Moller P.
        • Jacobsen N.R.
        • Folkmann J.K.
        • et al.
        Role of oxidative damage in toxicity of particulates.
        Free Radic Res. 2010; 44: 1-46
      1. Air Pollution and Cancer, IARC Scientific Publication No. 161, P8-P9.
      2. Cancer Epidemiology and Prevention. 3rd ed. Oxford University Press, Oxford, United Kingdom2006
        • World Health Organization
        Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks.
        World Health Organization, Geneva, Switzerland2009: 5-6
        • Turner M.C.
        • Krewski D.
        • Pope 3rd, C.A.
        • Chen Y.
        • Gapstur S.M.
        • Thun M.J.
        Long-term ambient fine particulate matter air pollution and lung cancer in a large cohort of never-smokers.
        Am J Respir Crit Care Med. 2011; 184: 1374-1381
        • Tseng C.H.
        • Chiang C.J.
        • Tseng J.S.
        • et al.
        EGFR mutation, smoking, and gender in advanced lung adenocarcinoma.
        Oncotarget. 2017; 8: 98384-98393
        • Liu M.
        • Bi J.
        • Ma Z.
        Visibility-based PM2.5 concentrations in China: 1957–1964 and 1973–2014.
        Environ Sci Technol. 2017; 51: 13161-13169

      Linked Article

      • Air Pollution and Adenocarcinoma in Never-Smokers
        Journal of Thoracic OncologyVol. 14Issue 5
        • Preview
          Exposure to atmospheric fine particles smaller than 2.5 μm (PM2.5) was recently estimated to have contributed 223,000 deaths from lung cancer worldwide in 2010.1 More than half of the lung cancer deaths attributable to ambient PM2.5 were estimated to have been in China and other East Asian countries. In a U.S. study, long-term exposure to PM2.5 was associated with lung cancer mortality.2,3 Mortality rates among male and female never-smokers were estimated at 17.1 and 14.7 per 100,000 person-years, respectively.
        • Full-Text
        • PDF
        Open Archive