The effects of pulmonary diseases on histologic types of lung cancer in both sexes: a population-based study in Taiwan

Background The associations between pulmonary diseases (asthma, chronic obstructive pulmonary disease [COPD], and tuberculosis [TB]) and subsequent lung cancer risk have been reported, but few studies have investigated the association with different histologic types of lung cancer. Methods Patients newly diagnosed with lung cancer from 2004 to 2008 were identified from the National Health Insurance Research Database in Taiwan. Histologic types of lung cancer were further confirmed using the Taiwan Cancer Registry Database. Cox proportional hazards regression was used to calculate the hazard ratio (HR) of asthma, COPD, and TB and to estimate the risk of specific types of lung cancer. Results During the study period, 32,759 cases of lung cancer were identified from 15,219,024 insurants aged 20 years and older. In men and women, the adjusted HR estimates of squamous cell carcinoma were respectively 1.37 (95 % confidence interval [CI], 1.21–1.54) and 2.10 (95 % CI, 1.36–3.23) for TB, 1.52 (95 % CI, 1.42–1.64) and 1.50 (95 % CI, 1.21–1.85) for asthma, and 1.66 (95 % CI, 1.56–1.76) and 1.44 (95 % CI, 1.19–1.74) for COPD. Similarly, the adjusted HR estimates of adenocarcinoma were respectively 1.33 (95 % CI, 1.19–1.50) and 1.86 (95 % CI, 1.57–2.19) for TB, 1.13 (95 % CI, 1.05–1.21) and 1.18 (95 % CI, 1.09–1.28) for asthma, and 1.50 (95 % CI, 1.42–1.59) and 1.33 (95 % CI, 1.25–1.42) for COPD. The HRs of small cell carcinoma were respectively 1.24 (95 % CI, 1.01–1.52) and 2.23 (95 % CI, 1.17–4.25) for TB, 1.51 (95 % CI, 1.35–1.69) and 1.63 (95 % CI, 1.16–2.27) for asthma, and 1.39 (95 % CI, 1.26–1.53) and 1.78 (95 % CI, 1.33–2.39) for COPD. Conclusions Asthma, COPD, and TB were associated with an increased risk of all major subtypes of lung cancer. The risk was the highest among women with TB.


Background
Lung cancer is the second leading diseases contributing to years of life lost because of premature mortality [1]. Among histologic types of lung cancer, adenocarcinoma is the most common subtype in Asians but not in Europeans [2,3]. Typical risk factors for lung cancer include smoking and exposure to arsenic, chromium, radon, or air pollution [4,5]. Smoking is the major risk factor for lung cancer, particularly squamous cell carcinoma (SqCC) [6]. However, a previous study demonstrated that most Taiwanese women with lung cancer are non-smokers [7]. A vast majority of smokers do not seem to develop lung cancer. Although smoking is a potential risk factor, other factors may also be linked to the increased risk of lung cancer.
Recent studies have concluded that chronic inflammation may be linked to lung carcinogenesis [8]. Among intrinsic pulmonary diseases, chronic obstructive pulmonary disease (COPD) [9,10], asthma [11], and tuberculosis (TB) [12] are associated with lung cancer. Smokers with COPD have a higher risk of SqCC [13,14]. Asthma is associated with an increased risk of SqCC and small cell carcinoma (SmCC) but is weakly associated with adenocarcinoma [15,16]. TB is also associated with an increased risk of SqCC and adenocarcinoma but not SmCC [17]. An association between TB and lung adenocarcinoma has been reported in non-westernized countries [18]. Furthermore, the association of lung cancer with diabetes [19] and dyslipidemia [20][21][22] has been reported. Data on pulmonary diseases and specific histologic types of lung cancer are considerable limited in Taiwan.
For a detailed evaluation of the relationship between pulmonary diseases and histologic types of lung cancer, a population-based cohort study is highly desirable. However, few such studies have been conducted. This study assessed whether pulmonary diseases are associated with an increased risk of specific types of lung cancer.

Database
The National Health Insurance Research Database (NHIRD) contains enrollment files, claims data, catastrophic illness files, and treatment registries. The national health insurance program covers more than 99 % of the population of Taiwan. The NHIRD is one of the largest administrative health care databases that is broadly used in academic studies [23][24][25]. This study used the linked databases of the NHIRD, Taiwan Cancer Registry Database (TCRD), and National Death Registry Database (NDRD) with the permission of the Department of Statistics, Ministry of Health and Welfare of Taiwan. The source data was encrypted and the data extracted was anonymous. This study was approved by the Institutional Review Board of the Chung-Shan Medical University Hospital, Taiwan.

Identification of Patients With Lung Cancer
In this study, 17,859,318 residents aged 20 years and older were initially enrolled. We excluded patients diagnosed with lung cancer before 2003 (n = 39,623) and those with incomplete information on sex (n = 2,600,565), registry data (n = 5), and death (n = 101). Finally, 15,219,024 patients (8,002,536 men and 7,216,488 women) were enrolled in this study. Patients newly diagnosed with lung cancer in 2004 were followed up until death, loss to follow-up, or the study end in 2008. Lung cancer was identified using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 162.
Furthermore, histologic types of lung cancer were confirmed using the TCRD. The registry contains data on cancer types, initial tumor stages, and histology. Lung cancer was identified using the ICD-9-CM code 162 or ICD 10 codes C34.0, C34. The linked databases were used to retrieve information on the age of lung cancer diagnosis, follow-up time (in person-years), and survival time and to minimize potentially unconfirmed cancer diagnoses.

Discussion
This study demonstrated that male and female patients with TB, asthma, and COPD had increased risks of lung SqCC, adenocarcinoma, and SmCC. Determining risk factors for specific types of lung cancer can help physicians gain a detailed understanding of the etiology of lung cancer and therefore identify the high-risk population for screening. To the best of our knowledge, no study has investigated the association between pulmonary diseases and histologic types of lung cancer. According to the study results, heterogeneity was observed in the risk factors for lung cancer and the different histologic types in male and female patients. A population-based, case-control study of female nonsmokers revealed an increased risk of lung cancer in patients with TB who were diagnosed before the age of 21 years [29]. The incidence rate ratio of lung cancer in the TB cohorts was 1.98 (95%CI, 1.37-2.83) 2 -4 years after TB infection [12].  [18]. A study conducted in Taiwan demonstrated that TB was an independent risk factor for SqCC, SmCC, and adenocarcinoma in men and women [6]. Such an association is particularly crucial in Taiwan, where the prevalence of TB is high [31,32]. Compared with asthma and COPD, TB appears to have a stronger association with lung cancer among women. Additional studies are necessary to assess the possible mechanisms of this association.
In this study, COPD was associated with the risk of the major types of lung cancer. Chronic airway inflammation is a major risk factor for COPD and is also associated with an increased risk of lung cancer [13]. A study involving the 22-year follow-up of 5,402  [34]. The prevalence of smoking is almost 10fold higher in Taiwanese men than that in women [35]. However, the smoking status of the study population was not available. This may be the reason for the observed differences between men and women. Chronic bronchitis and emphysema increased the risk of SqCC (HR, 1.54; 95 % CI, 1.09-2.18) independent of smoking [36]. COPD also increased the risk of SqCC in smokers [13]. Pesch et al. performed a pooled analysis of case-control studies including 13,169 cases and 16,010 controls from Europe and Canada [37]. Their analysis demonstrated that adenocarcinoma was the most prevalent subtype in never-smokers and women. The ORs were elevated for exposure to cigarette smoke and were higher for SqCC and SmCC than for adenocarcinoma. Freedman et al. recruited 279,214 men and 184,623 women aged 50-71 years from eight states in the United States to evaluate whether women were more susceptible to lung cancer caused by cigarette smoking than men [38]. Their results illustrated that the HRs of adenocarcinoma, SmCC, and undifferentiated tumors were similar between men and women among ex-smokers and current smokers. However, among current smokers, the HR of SqCC in men was approximately 2-fold higher than that in women. This study also suggests that patients with asthma are at an increased risk of three histological types (SqCC, adenocarcinoma, and SmCC) of lung cancer. Asthma is one of the most common chronic airway diseases and affects 300 million people of all ages and ethnicities [39]. In Taiwan, the prevalence of asthma has increased to 11.9 % [40]. Because asthma causes complex chronic airway inflammation, it has been hypothesized to lead to carcinogenesis [8]. Case-control studies have produced varied results for the association between asthma and lung cancer [41,42]. In a Swedish cohort with a hospital-discharge diagnosis of asthma, the standardized incidence rate ratio of lung cancer was 1.51 in men (95 % CI, 1.38-1.65) and 1.78 in women (95 % CI, 1.55-2.03), and the risk of histologic types of lung cancer was higher in patients with SqCC and SmCC [16]. In a meta-analysis, the RRs were 1 [15].
In this study, hyperlipidemia was also associated with an increased risk of adenocarcinoma. Hyperlipidemia is a component of metabolic syndrome and is associated Smoking-related cancers included lip, oral cavity, nasal cavity, pharynx, larynx, and esophagus, pancreas, kidney and bladder cancers that were prior to a diagnosis of lung cancer Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; HR, hazard ratio with insulin resistance [43]. Hyperinsulinemia, hyperglycemia, and chronic inflammation play a vital role in the neoplastic process [44]. High serum triglyceride concentrations are associated with an increased risk of lung cancer (fourth vs first quartile: HR, 1.94; 95 % CI, 1.47-2.54) [22]. Additional studies are required to assess the association between hyperlipidemia and adenocarcinoma. Evaluating the temporal relationship between pulmonary diseases and subsequent lung cancer in case-control studies is difficult. Previous studies might have yielded inconclusive results because they focused mainly on the high-risk populations of heavy smokers. This study has several strengths. First, our data were retrieved from combined databases (NHIRD, TCRD and NDRD) that included all residents; hence, recall and selection bias was minimized. Second, the histologic type of lung cancer was confirmed using the TCRD. Nevertheless, our study has some limitations. First, the NHIRD does not contain detailed clinical data and information on lifestyle-related factors such as smoking, obesity, physical inactivity, dietary habits, and family history, which are closely associated with lung cancer. Smoking is a major confounding factor of lung cancer. Biases were minimized by adjusting for COPD and smoking-related cancer. Second, patients with asthma, COPD, and TB may have used medications that may have complicated their conditions. This study did not evaluate the effects of drugs.

Conclusions
This study demonstrated that asthma, COPD, and TB were associated with increased risks of all major subtypes of lung cancer. The risk was the highest among women with TB.

Competing Interests
The authors declare that they have no competing interests.
Author Contributions JYH, ZHJ, and YPL conceived and designed the study. JYH, PCK, and CCL performed the experiments. ZHJ, JYH, WYK, CCH, and HHP analyzed the data. CYH, YCL, and CCL contributed analysis tools. ONN, CCL, WYK, and CCH provided critical inputs on design, analysis, and interpretation of the study. All the authors had access to the data. All authors read and approved the final manuscript as submitted. Smoking-related cancers included lip, oral cavity, nasal cavity, pharynx, larynx, and esophagus, pancreas, kidney and bladder cancers that were prior to a diagnosis of lung cancer Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; HR, hazard ratio