Oxidative damage by cigarette smoking-induced free radicals is thought to play an important role in the development of lung cancer. Using urine from a prospective cohort of more than 18,000 men, we quantified 8-epi-prostaglandin F20 in baseline urine. Smokers with elevated levels 8-epi-prostaglandin F20doubled risk of lung cancer. A heightened state of oxidative stress in response to the insults of cigarette smoking contributes to lung carcinogenesis.
Abstract
It has been hypothesized that the pathogenesis of lung cancer induced by cigarette smoking involves oxidative damage by free radicals. Epidemiological data on biomarkers of oxidative damage and risk of lung cancer development are sparse. A nested case–control study of 610 lung cancer cases and 610 matched controls was conducted within a prospective cohort of 18 244 Chinese men in Shanghai, China. The concentrations of 8-epi-prostaglandin F2α (8-epiPGF2α), a biomarker of oxidative stress, were determined in baseline urine samples using a validated mass-spectrometry assay. Current smokers had significantly higher level of 8-epiPGF2α than former smokers or never smokers (P < 0.001). 8-epiPGF2α levels were significantly higher in lung cancer cases than their smoking-matched controls in former and current smokers, but not different in never smokers (P for interaction = 0.019). The relative risks of developing lung cancer for former and current smokers in the highest relative to the lowest quartile of 8-epiPGF2α were 5.25 (Ptrend = 0.035) and 1.99 (Ptrend =0.007), respectively. The effect of 8-epiPGF2α and biomarkers of cigarette smoke exposure on lung cancer risk was additive; the relative risk was 5.33 (95% confidence interval = 2.65–7.51) for current smokers with the highest thirds of 8-epiPGF2α and total cotinine compared with their lowest thirds. Smokers with a heightened state of oxidative stress in response to the insults of cigarette smoking may be more susceptible to smoking-induced lung carcinogenesis.
Introduction
Lung cancer is the leading cause of cancer death in both men and women worldwide (1.5 million deaths per year) and in the USA (158 000 deaths per year) (1,2). Tobacco smoking is the most important causal factor for lung cancer. It is estimated that 90% of all lung cancer deaths in the USA are attributable to cigarette smoking, while it accounts for 80% of the worldwide lung cancer burden in males and at least 50% in females (3,4). However, there is considerable variation in susceptibility to lung cancer among smokers; only 24% of male lifelong smokers and 12% of female lifelong smokers die from lung cancer by 85 years of age after taking into account competing risks of death (5). This large inter-individual variation in smoking-related lung cancer risk may be determined in part by variability in the uptake and metabolism of tobacco smoke carcinogens. There are 72 established carcinogens in cigarette smoke (6). Among these, polycyclic aromatic hydrocarbons and the tobacco-specific nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) are widely considered to be among the most important causative agents for lung cancer (7,8). We previously reported that the urinary levels of total 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), a metabolite of NNK, phenanthrene tetraol (PheT), a biomarker of polycyclic aromatic hydrocarbon and total cotinine, a metabolite of nicotine and biomarker of overall tobacco constituent uptake, were associated with an increased risk of lung cancer in two prospective cohorts of Chinese smokers (9,10).
Decades of research have firmly established that cigarette smoke and its condensate have tumor promoting and co-carcinogenic activity, which collaborate with the multiple carcinogens in smoke to cause cancer. While the mechanisms by which these tumor promoting and co-carcinogenic agents act are complex and not fully understood, there is little doubt that oxidative damage and inflammation play a significant role (11–17). Epidemiologic studies of lung cancer have not previously investigated this combination using appropriate biomarkers reflecting both carcinogen exposure and oxidative stress.
8-epi-Prostaglandin F2α (8-epiPGF2α) is a product of lipid peroxidation and a reliable biomarker of in vivo oxidative stress (18,19). 8-epiPGF2α is excreted in urine and chemically stable (20). Urinary levels of 8-epiPGF2α are elevated in smokers, and are associated with number of cigarettes smoked per day (21). Epidemiological studies on the relationship of 8-epiPGF2α to the risk of lung cancer are sparse. A cross-sectional study reported that serum levels of 8-isoprostane were elevated in lung cancer patients than control subjects, and were higher in patients with advanced disease (22). A nested case–control study within the Multiethnic Cohort found that baseline urinary levels of 15-isoprostane F2α (the same as 8-epiPGF2α) were related to lung cancer risk in men, but not in women (23).
Utilizing the resources of the Shanghai Cohort Study, the present study examined the association between urinary levels of 8-epiPGF2α and risk of lung cancer in all subjects, and in never, former and current smokers separately. Among current smokers, we further assessed the independent association between urinary levels of 8-epiPGF2α and lung cancer risk with adjustment for biomarkers of cigarette smoke exposure in addition to self-reported smoking history.
Material and methods
Study population
Subjects were drawn from the Shanghai Cohort Study (24). Briefly, the Shanghai Cohort consisted of 18 244 men enrolled from 1 January 1986 through 30 September 1989, who were 45–64 years of age and resided in the city of Shanghai, China. In addition to in-person interviews eliciting information on use of tobacco and alcohol, a 10-ml blood sample and a single-void urine specimen were collected from each participant at baseline. The Shanghai Cohort Study has been approved by the Institutional Review Boards at the Shanghai Cancer Institute and the University of Pittsburgh. The present study has been approved by the Institutional Review Boards at the University of Minnesota and the University of Pittsburgh.
Nested case–control study
Identification of incident lung cancer cases was accomplished through annual in-person re-interviews of all surviving cohort members and routine review of reports from the population-based Shanghai Cancer Registry. To date, losses to follow-up totaled 839 individuals (4.6%) of the entire cohort after 25 years of study. The nested case–control design was previously described (10,25). As of 31 December 2006, 706 cohort participants developed lung cancer. Among them, 574 were current smokers, 43 were former smokers and 89 were never smokers at baseline. To increase sample sizes for never and former smokers, we included 42 never smokers and 16 former smokers who were diagnosed with lung cancer during 2007 through 2013. For each case, we randomly selected one control subject from all cohort members who were free of cancer and alive at the time of cancer diagnosis of the index case. Controls were matched to the index case by age at enrollment (±2 years), date of biospecimen collection (±1 month), neighborhood of residence and smoking status (never, former and current smokers) at recruitment.
Laboratory measurements
Urine samples of all study subjects were retrieved from the biospecimen bank. Specimens from matched control subjects and their index cases were always assayed in the same batch. All urine aliquots were identified only by unique codes, and laboratory personnel had no knowledge of the case/control status of the test samples. The assay for quantifying 8-epiPGF2α in urine was described previously (26). Briefly, urine (0.2 ml) was mixed with the internal standard [D4]8-epiPGF2α (2 ng) and the sample was acidified with 20 µl 88% formic acid, then applied to Agilent Bond Elut C18 solid phase extraction plates, previously conditioned with 1 ml methanol, 1 ml acetonitrile and 3 ml 50 mM phosphate buffer, pH 3. The plates were washed with 3 ml 50 mM phosphate buffer and 3 ml hexanes, then 8-epi-PGF2α was eluted with 1 ml ethyl acetate, collected in 96-well Tru-Taper plates, dried and stored at −20°C until analysis by LC-ESI-MS/MS. For the analysis, the samples were redissolved in 30 µl 80:20:25 H2O:MeOH:NH4OH and injected on a 50 × 1.0 mm Waters X-Bridge BEH C18 column with elution by a program from 5% B to 18% B in 11 min at a flow rate of 45 µL/min where solvent A was 0.15% NH4OH and Solvent B was 95:5:0.15 acetonitrile: methanol: NH4OH. LC-ESI-MS/MS was carried out on a Thermo Dionex Ultimate 3000 LC coupled to a Thermo Scientific TSQ Vantage triple quadrupole mass spectrometer, in the negative ESI mode, monitoring m/z 193 → m/z 173 for the analyte and m/z 197 → m/z 177 for the internal standard.
The measurements of serum concentrations of retinol, alpha- and gamma-tocopherols, and specific carotenoids including alpha-carotene, beta-carotene, beta-cryptoxanthin, lycopene and lutein/zeaxanthin by HPLC methods were described in our previous report (25). The measurements of urinary total NNAL, PheT and total cotinine were also previously reported (9,10). Urinary creatinine was assayed using a Kodak Ektachem 500 chemistry analyzer.
Of the 764 case–control pairs, 103 cases (92 current smokers and 11 never smokers) and 66 controls (60 current smokers, 1 former smoker and 5 never smokers) did not have urinary 8-epiPGF2α measurements because their urine samples collected at baseline were depleted after quantitation of other urinary biomarkers in previous studies (9,10,27–29). In addition, we excluded 51 cases without their matched controls and 88 controls without their index cases. The present study included the remaining 610 cases of lung cancer and 610 their individually matched control subjects.
Statistical analysis
Urinary 8-epiPGF2α was expressed in pg per mg creatinine (Cr) to correct for varying water contents of individual spot urine samples. Given the markedly skewed distributions of the urinary biomarkers, formal statistical testing was performed on logarithmically transformed values, and geometric means and 95% confidence intervals (CIs) are presented.
The χ2 test and the t-test were used to compare the distributions of selected variables between lung cancer cases and controls. The analysis of covariance method was used to examine the difference in the Cr-adjusted concentration of urinary 8-epiPGF2α across different smoking status/dose groups among control subjects as well as between lung cancer cases and controls with adjustment for potential confounders. The two-way analysis of variance method was used to simultaneously examine the effect of cigarette smoking and lung cancer status on levels of 8-epiPGF2α.
The relationship between urinary 8-epiPGF2α and lung cancer risk was assessed by odds ratios (ORs) and their 95% CIs and P values using the conditional logistic regression method that took into account for the effect of matching factors including age, year of biospecimens collection and neighborhood of residence at enrollment on lung cancer risk. In the preliminary analysis, we noticed that urinary levels of 8-epiPGF2α were significantly higher in current smokers than never and former smokers. Thus study subjects were grouped into quartiles according to the distributions of 8-epiPGF2α among separate groups of controls by smoking status. In the analysis for the association between 8-epiPGF2α and risk of lung cancer among current smokers, we further adjusted for biomarkers of exposure to cigarette smoke including total cotinine, total NNAL and PheT, the identified biomarkers that were independently associated with lung cancer in this cohort (9,10). Unconditional logistic regression models were used for all stratified analyses. An ordinal value of 8-epiPGF2α in quartile was used for all linear trend tests for its relation to lung cancer risk.
We examined the potential interaction effects between 8-epiPGF2α and levels of self-reported smoking history or biomarkers of cigarette smoke constituents on lung cancer risk. When examining whether the combined effect of two factors was greater than the multiplicative product of their individual effects, we used multivariate logistic regression models with the two main effects and their product term as covariates. When assessing if the combined effect of two factors on lung cancer risk was greater than the sum of the individual effects, we used the method described by Rothman (30). Under the null hypothesis of additivity, the synergy index would take on the value one. A lower 95% CI of the synergy index that is greater than one is indicative of a statistically significant synergistic effect between the two factors.
Statistical analyses were carried out using SAS software version 9.3 (SAS Institute, Cary, NC). All P-values reported are two-sided, and those that were less than 0.05 were considered to be statistically significant.
RESULTS
Of the 610 cases, 424 (69.5%) were histopathologically confirmed as lung cancer while the remaining 186 (30.5%) were diagnosed based on radiography or computer-assisted tomography evidence. Among the histopathologically confirmed cases, 171 (40.3%) were squamous cell cancers, 150 (35.4%) adenocarcinomas, 25 (5.9%) small cell cancers and 78 (18.4%) other cell types. The mean age (standard deviation) at cancer diagnosis of all case patients was 69.0 (7.6) years. The mean (SD) time interval between baseline biospecimen collection and cancer diagnosis was 11.3 (6.1) years, ranging from 1 month to 26.8 years.
Compared with controls, cases had lower body mass index (BMI) (kg/m2). Smoking status was matched between cases and controls. Among ever smokers, cases had greater numbers of cigarettes smoked per day, years of smoking and pack-years of smoking than controls. The percentage of regular drinkers of alcohol was comparable between cases and controls. Among drinkers of alcohol, cases consumed more drinks per day than their controls (Table 1). As previously reported (10), the geometric means of urinary total cotinine, total NNAL and PheT in lung cancer cases were significantly higher than those in controls among current smokers (Table 1).
Table 1.
Baseline demographic and lifestyle characteristics and urinary biomarkers of cigarette smoke exposure in lung cancer cases and matched control subjects, The Shanghai Cohort Study 1986–2013
| Characteristics or biomarkers | Cases | Controls | P* |
|---|---|---|---|
| Number of subjects | 610 | 610 | |
| Mean age (SD), years | 57.3 (5.2) | 57.1 (5.0) | 0.513a |
| Men, N (%) | 610 (100) | 610 (100) | 1.000a |
| Mean body mass index (SD), kg/m2 | 21.6 (2.8) | 22.1 (3.1) | 0.001 |
| Level of education, N (%) | |||
| No formal education | 51 (8.4) | 41 (6.7) | 0.225 |
| Primary school (1–6 years) | 204 (33.4) | 186 (30.5) | |
| Secondary school or higher | 355 (58.2) | 383 (62.8) | |
| Smoking status, N (%) | |||
| Non-smokers | 116 (19.0) | 116 (19.0) | 1.000a |
| Former smokers | 58 (9.5) | 58 (9.5) | |
| Current smokers | 436 (71.5) | 436 (71.5) | |
| Mean cigarettes/day (SD)b | 20.2 (8.6) | 15.4 (7.6) | <0.001 |
| Mean years of smoking (SD)b | 34.7 (8.5) | 30.6 (10.5) | <0.001 |
| Mean pack-years of smoking (SD)b | 35.5 (18.0) | 24.3 (15.3) | <0.001 |
| Alcohol drinking status, N (%) | |||
| Non-drinkers | 292 (47.9) | 278 (45.6) | 0.421 |
| Regular drinkers | 318 (52.1) | 332 (54.4) | |
| Mean drinks/day (SD)c | 3.0 (2.7) | 2.5 (2.2) | 0.005 |
| Urinary biomarkers of cigarette smoking | |||
| Total cotinine (nmol/mg Cr)d | 13.38 (12.16, 14.76) | 6.98 (6.32, 7.68) | <0.001 |
| Total NNAL (pmol/mg Cr)d | 0.28 (0.26, 0.30) | 0.20 (0.18, 0.20) | <0.001 |
| PheT (pmol/mg Cr)d | 32.38 (30.66, 34.18) | 28.10 (26.62, 29.68) | <0.001 |
Cr, creatinine; SD, standard deviation.aAge, sex and smoking status were matched between cases and controls.
bAmong ever smokers.
cAmong regular alcohol drinkers only.
dAmong current smokers only; geometric means (95% confidence intervals) are shown.
*Two-sided P’s were based on t test for continuous variables or chi-square test for categorical or nominal variables.
Cigarette smoking had a significant impact on urinary levels of 8-epiPGF2α. In both lung cancer cases and controls, current smokers had significantly higher levels of urinary 8-epiPGF2α than never and former smokers (Table 2). Among controls of current smokers, levels of 8-epiPGF2α were positively associated with number of cigarettes per day (ρ = 0.20, P < 0.001), urinary total cotinine (ρ = 0.33, P < 0.001), urinary total NNAL (ρ = 0.26, P < 0.001) and urinary PheT (ρ = 0.20, P < 0.001), but not associated with years of smoking (ρ = 0.02, P =0.687). Among former smokers, the 8-epiPGF2α levels were not significantly correlated with number of years since quitting smoking in cases (ρ = −0.14, P =0.311) and controls (ρ = −0.20, P =0.133), respectively. Among all controls, BMI had a positive association with 8-epiPGF2α in former smokers (ρ = 0.22, P = 0.098), and an inverse association in current smokers (ρ = −0.21, P <0.001), but no association in never smokers (ρ = 0.008, P = 0.934). Of the 610 control subjects, 128 had serum measurements of antioxidants from a previous study (25). Urinary levels of 8-epiPGF2α were significantly or borderline significantly associated with lower levels of serum α-tocopherol (ρ = −0.21, P = 0.018), serum α-carotene (ρ = −0.16, P = 0.069) and serum lycopene [(ρ = −0.18, P = 0.049), but not associated with other serum antioxidants including retinol, β-carotene, β-cryptoxanthin or lutein/zeaxanthin (data not shown)].
Table 2.
Geometric means of 8-epi-prostaglandin F2α (8-epiPGF2α) in lung cancer cases and control subjects by smoking status, The Shanghai Cohort Study 1986–2013
| Subject group | Cases | Controls | P for cases versus controls | ||
|---|---|---|---|---|---|
| N | 8-epiPGF2αa (pg/mg creatinine) |
N | 8-epiPGF2αa (pg/mg creatinine) |
||
| Total subjects | 610 | 456 (434–479) | 610 | 402 (382–422) | <0.001 |
| By smoking status | |||||
| Never smokers | 116 | 248 (226–273) | 116 | 254 (231–280) | 0.743 |
| Former smokers | 58 | 293 (256–335) | 58 | 222 (194–255) | 0.005 |
| Current smokers | 436 | 573 (545–602) | 436 | 488 (464–512) | <0.001 |
| P for current versus never smokers | <0.001 | <0.001 | P for interaction = 0.019† | ||
| P for current versus former smokers | <0.001 | <0.001 | |||
| P for former versus never smokers | 0.113 | 0.051 | |||
aGeometric means (95% confidence intervals) and P values were derived from the two-way analysis of variance with log (8-epiPGF2α) as dependent variable and lung cancer status and smoking status as the two main factors with the following covariates: age, year of biospecimen collection, neighborhood of residence at baseline, level of education and body mass index (kg/m2).
†Two-sided P for interaction between smoking status (current versus never) and lung cancer status on log (8-epiPGF2α).
The geometric mean of creatinine-corrected 8-epiPGF2α in lung cancer cases was significantly higher than that in controls after adjustment for age, level of education, BMI, smoking status, number of cigarettes per day and years of smoking (Table 2). The differences in geometric means of 8-epiPGF2α between cases and controls within each category of smoking status were different, with the largest case–control difference being seen in current smokers (P < 0.001), a moderate difference in former smokers (P = 0.005), but no difference in never smokers (P = 0.743). The interaction between smoking status (current versus never) and lung cancer status on urinary concentrations of 8-epiPGF2α was statistically significant (P for interaction = 0.019) after adjustment for multiple potential confounding factors (Table 2).
Higher levels of urinary 8-epiPGF2α were associated with a higher risk of lung cancer. Among all subjects, the risk of developing lung cancer was doubled for individuals in the highest quartile of 8-epiPGF2α relative to the lowest quartile after adjustment for history of smoking and other potential confounders (P for trend = 0.007) (Table 3). This positive association was present in both current and former smokers. Compared with the lowest quartile of 8-epiPGF2α, the multivariable-adjusted ORs of developing lung cancer for the highest quartile in former and current smokers were 5.25 (95% CI = 0.94–29.41, P for trend = 0.035) and 1.99 (95% CI = 1.26–3.14, P for trend = 0.007), respectively. However, there was no positive relationship of 8-epiPGF2α to lung cancer risk among never smokers (P for trend = 0.538).
Table 3.
Urinary levels of 8-epi-prostaglandin F2α (8-epiPGF2α) in relation to lung cancer risk in total and subgroups by smoking status, The Shanghai Cohort Study 1986–2013
| 8-epiPGF2α in quartile (pg/mg Cr)a | Cases | Controls | OR (95% CI)b |
|---|---|---|---|
| All subjects | |||
| First quartileb | 93 | 150 | 1.00 |
| Second quartile | 139 | 151 | 1.31 (0.88, 1.94) |
| Third quartile | 147 | 156 | 1.18 (0.80, 1.74) |
| Fourth quartile | 231 | 153 | 1.79 (1.22, 2.63) |
| P for trend | 0.007 | ||
| Never smokers | |||
| First quartile (<172) | 23 | 29 | 1.00 |
| Second quartile (172–241) | 32 | 27 | 1.62 (0.70, 3.78) |
| Third quartile (242–338) | 37 | 30 | 1.54 (0.68, 3.52) |
| Fourth quartile (>338) | 24 | 30 | 0.71 (0.29, 1.75) |
| P for trend | 0.528 | ||
| Former smokers | |||
| First quartile (<165) | 11 | 15 | 1.00 |
| Second quartile (165–210) | 7 | 14 | 1.04 (0.19, 5.75) |
| Third quartile (211–267) | 8 | 15 | 0.83 (0.21, 3.34) |
| Fourth quartile (>267) | 32 | 14 | 5.25 (0.94, 29.41) |
| P for trend | 0.035 | ||
| Current smokers | |||
| First quartile (<340) | 59 | 106 | 1.00 |
| Second quartile (340–487) | 100 | 110 | 1.33 (0.83, 2.12) |
| Third quartile (488–650) | 102 | 111 | 1.13 (0.70, 1.82) |
| Fourth quartile (>650) | 175 | 109 | 1.99 (1.26, 3.14) |
| P for trend | 0.007 | ||
aQuartile cut-off values were based on the distribution of urinary 8-epiPGF2α among control subjects stratified by smoking status.
bOdds ratios (95% confidence intervals) were derived from conditional logistic regression models with adjustment for level of education, body mass index (kg/m2), smoking status, number of cigarettes/day, years of smoking and years since quitting smoking (for former smokers only).
Among former smokers, more than half of cases but less than a quarter of controls were in the highest quartile of 8-epiPGF2α at baseline, yielding an OR of 5.29 (95% CI = 1.58–17.72), comparing with the first, second and third quartiles combined. Among the 32 cases of former smokers with the highest quartile of 8-epiPGF2α, the median time interval from quitting smoking to the study enrollment (i.e. collection of biospecimen) was 15.6 years. Thus, the elevation of urinary 8-epiPGF2α for these cases was less likely related to recent smoking cessation.
Among current smokers, additional adjustment for biomarkers of cigarette smoke exposure including total cotinine, total NNAL and PheT, diminished the OR of lung cancer associated with higher levels of 8-epiPGF2α. Compared with the lowest quartile of 8-epiPGF2α, the biomarker-adjusted OR of lung cancer for the highest quartile was 1.44 (95% CI = 0.88–2.36) (data not shown). We further analyzed for the association for the joint levels of 8-epiPGF2α and these biomarkers of cigarette smoke exposure with lung cancer risk among current smokers. Compared with the lowest tertiles of both urinary 8-epiPGF2α and total cotinine, the OR of developing lung cancer for the highest tertiles of both 8-epiPGF2α and total cotinine was 5.33 (95% CI = 2.64–7.51) (Table 4). Similarly, ORs (95% CIs) for 8-epiPGF2α in combination with total NNAL and PheT were 3.84 (95% CI = 1.93–7.65) and 2.90 (95% CI = 1.42–5.92), respectively, after adjustment for smoking intensity and duration and other potential confounders (Table 4). However, these combined effects were not more than the sum of the two individual effect of 8-epiPGF2α and total cotinine (or other biomarkers of cigarette smoking) on the estimates of lung cancer risk.
Table 4.
Joint tertile levels of urinary 8-epi-prostagla ndin F2α with urinary biomarkers of cigarette smoke exposure in relation to risk of lung cancer among current smokers, The Shanghai Cohort Study 1986–2013
| Biomarkers of cigarette smoke exposure | Levels in tertile | Urinary 8-epi-prostaglandin F2α in tertile | |||||
|---|---|---|---|---|---|---|---|
| First tertile | Second tertile | Third tertile | |||||
| Ca/Coa | OR (95% CI)b | Ca/Coa | OR (95% CI)b | Ca/Coa | OR (95% CI)b | ||
| Total cotinine | First tertile | 15/75 | 1.00 (ref) | 13/43 | 1.15 (0.43, 3.03) | 14/28 | 1.95 (0.75, 5.08) |
| Second tertile | 32/37 | 3.70 (1.65, 8.28) | 37/61 | 1.77 (1.65, 8.28) | 52/45 | 3.32 (1.56, 7.04) | |
| Third tertile | 42/29 | 2.93 (1.33, 6.42) | 71/47 | 3.55 (1.68, 7.51) | 160/71 | 5.33 (2.64, 7.51) | |
| Total NNAL | First tertile | 21/60 | 1.00 (ref) | 17/51 | 0.78 (0.35, 1.76) | 21/35 | 1.24 (0.53, 2.89) |
| Second tertile | 37/52 | 1.52 (0.72, 3.21) | 47/52 | 1.37 (0.65, 2.91) | 64/44 | 2.10 (1.05, 4.19) | |
| Third tertile | 31/29 | 1.95 (0.84, 4.53) | 57/47 | 2.03 (0.96, 4.28) | 141/65 | 3.84 (1.93, 7.65) | |
| PheT | First tertile | 29/57 | 1.00 (ref) | 28/59 | 0.67 (0.30, 1.46) | 48/28 | 2.61 (1.19, 5.75) |
| Second tertile | 26/45 | 1.30 (0.61, 2.79) | 43/50 | 1.59 (0.77, 3.28) | 67/49 | 2.16 (1.04, 4.49) | |
| Third tertile | 34/39 | 1.67 (0.72, 3.90) | 50/42 | 2.05 (0.93, 4.53) | 111/67 | 2.90 (1.42, 5.92) | |
aCa/Co: number of case and control.
bOdds ratios (95% confidence intervals) were derived from conditional logistic regression models with adjustment for level of education, body mass index (kg/m2), number of cigarettes per day and years of smoking.
We also examined association between the levels of 8-epiPGF2α and risk of lung cancer among current smokers stratified by levels of cigarette smoking. The association between 8-epiPGF2α and lung cancer risk was comparable for <20 cigarettes/day with ≥20 cigarettes/day (P for interaction = 0.203) or for smokers with different pack-years of smoking—<20, 20–39 and ≥40 pack-years of smoking (P for interaction = 0.722). The duration of smoking had a borderline significant modifying effect on the association between 8-epiPGF2α and lung cancer risk (P for interaction = 0.056) (Supplementary Table 1, available at Carcinogenesis Online).
The potential impact of disease progression on 8-epiPGF2α was examined among 610 cases by the time interval from biospecimen collection to lung cancer diagnosis. Urinary level of 8-epiPGF2α was not correlated with the time interval (ρ = −0.007, P = 0.859). The geometric means of 8-epiPGF2α were similar for patients whose urine samples for 8-epiPGF2α were collected within 2 years as for those collected more than 20 years before cancer diagnosis in all cases (Supplementary Table 2, available at Carcinogenesis Online). The associations between quartile levels of urinary 8-epiPGF2α and lung cancer risk among ever smokers did not attenuate for longer than shorter duration of follow-up from biospecimens collection to lung cancer diagnosis. A statistically significant, positive association for urinary 8-epiPGF2α was still present for lung cancer cases diagnosed 10 or more years after the collection of urine samples for measurement of 8-epiPGF2α (Table 5).
Table 5.
Urinary levels of 8-epi-prostaglandin F2α (8-epiPGF2α) in relation to lung cancer risk by years between biospecimen collection and cancer diagnosis among ever smokers, The Shanghai Cohort Study 1986–2013
| 8-epiPGF2α | Controls | <5 years | 5–<10 years | 10–<15 years | ≥15 years | ||||
|---|---|---|---|---|---|---|---|---|---|
| Cases | OR (95% CI)a | Cases | OR (95% CI)a | Cases | OR (95% CI)a | Cases | OR (95% CI)a | ||
| First quartile | 121 | 15 | 1.00 | 17 | 1.00 | 15 | 1.00 | 23 | 1.00 |
| Second quartile | 124 | 27 | 1.59 (0.77, 3.25) | 29 | 1.82 (0.90, 3.68) | 27 | 1.37 (0.67, 2.77) | 24 | 1.04 (0.54, 2.00) |
| Third quartile | 126 | 23 | 1.30 (0.62, 2.72) | 31 | 1.58 (0.79, 3.18) | 27 | 1.27 (0.62, 2.58) | 29 | 1.11 (0.58, 2.10) |
| Fourth quartile | 123 | 35 | 1.96 (0.97, 3.98) | 58 | 2.70 (1.39, 5.24) | 57 | 2.73 (1.40, 5.33) | 57 | 1.88 (1.04, 3.40) |
| P for trend | 0.109 | 0.006 | 0.002 | 0.024 | |||||
aOdds ratios (95% confidence intervals) were derived from unconditional logistic regression models with adjustment for matching factors (age, years of biospecimen collection, neighborhood of residence at baseline), level of education, body mass index (kg/m2), smoking status, number of cigarettes/day and years of smoking.
The associations between urinary 8-epiPGF2α and risk of lung cancer differed by histological subtype. The levels of 8-epiPGF2α were significantly associated with increased risk lung squamous cell carcinoma (P for trend = 0.016), but not with risk of lung adenocarcinoma (P for trend = 0.354) (Supplementary Table 3, available at Carcinogenesis Online). However, the difference between the two risk associations was not statistically significant (P = 0.194).
Of the 610 lung cancer cases and 610 control subjects included in the present study, 182 cases and 128 controls had measurements of serum antioxidants from our previous study (25). In this subset, ORs of lung cancer for highest relative to the lowest quartile of urinary 8-epiPGF2α were 1.63 (95% CI = 0.74–3.59, P for trend = 0.202) after adjustment for matching factors (age, year of biospecimens collection and neighborhood of residence), level of education, BMI and history of smoking (smoking status, number of cigarettes/day, years of smoking and years since quitting smoking for former smokers). Further adjustment for serum antioxidants including α- and β-carotenes, β-cryptoxanthin, lycopene, lutein/zeaxanthin, α- and γ-tocopherols and retinol did not materially change the association between 8-epiPGF2α and risk of lung cancer; the antioxidants-adjusted OR was 1.79 (95% CI = 0.78–4.11, P for trend = 0.163).
Discussion
In this nested case–control study within the Shanghai Cohort Study, we found a statistically significant, independent association between the urinary level of 8-epiPGF2α and risk of lung cancer after adjustment for smoking intensity and duration and other potential confounding factors. The overall risk of lung cancer development was doubled for subjects in the highest quartile of 8-epiPGF2α. The positive association remained statistically significant even after further adjustment for multiple biomarkers of cigarette smoke exposure and serum antioxidants. The significant association between urinary 8-epiPGF2α and lung cancer risk was present in both former and current smokers. The number of years since quitting smoking did not confound the 8-epiPGF2α-lung cancer risk association among former smokers. For current smokers, there was evidence supporting an additive effect of 8-epiPGF2α and biomarkers of cigarette smoke constituents including total cotinine, total NNAL and PheT. This is the first prospective study that examined the joint effect of 8-epiPGF2α and these biomarkers of cigarette smoking on lung cancer risk.
8-epiPGF2α is a reliable marker and considered a gold standard for oxidative damage via lipid peroxidation. Several epidemiological studies have reported positive associations between urinary 8-iso-PGF2α levels and risk of multiple cancers including lung cancer (23), breast cancer (31), liver cancer (32) and prostate cancer (33). In the only reported study on lung cancer, higher levels of 15-isoprostane F2t (the same as 8-epiPGF2α) in prediagnostic urine samples were associated with a statistically non-significant, 2-fold risk of lung cancer in men (P for trend = 0.12), but were not associated with lung cancer risk in women (23). The non-significant results could be due to the relatively small sample size of the study that included only 136 male cases and 71 female cases. Another limitation of that study was the use of an enzyme-linked immunosorbent assay (ELISA) to quantify urinary 15-F2t-isoP, which may have cross-reactivity with related compounds and result in low specificity and an overestimation (26). In this study, we used the LC-MS/MS method that provides high specificity, accuracy and precision (26). Furthermore, other biomarkers of cigarette smoke exposure including total cotinine, total NNAL and PheT were measured on all study subjects from a previous study (10), allowing for investigating the independent effect and joint effect of 8-epiPGF2α on lung cancer risk.
The increased risk of lung cancer with high levels of 8-epiPGF2α was present in former and current smokers, but not in never smokers. Urinary levels of 8-iso-PGF2α were significantly higher in current smokers than never and former smokers in the present study populations. These findings are similar to those of previous studies (21,26,34–36). In both former and current smokers, urinary 8-epiPGF2α levels at baseline, on average, 11 years before cancer diagnosis, were significantly higher than their controls. These data are fully consistent with previous mechanistic studies of tobacco carcinogenesis which clearly demonstrate that tumor promotion and co-carcinogenesis, mediated by oxidative damage and inflammation, are critical events in cancer development caused by cigarette smoking (11–17). Thus, this is the first study to demonstrate the joint effect of carcinogen exposure and oxidative damage in increasing the risk for lung cancer in cigarette smokers. The consistency of our data on exposure biomarkers representing tobacco-specific nitrosamines and polycyclic aromatic hydrocarbon and effect biomarkers such as 8-epi-PGF2a is remarkable and supports the multiple mechanistic studies that implicate these factors in tobacco carcinogenesis.
The present study has several strengths. First, the LC-MS/MS method for quantification of 8-epiPGF2α provides high quality data. Second, we used a comprehensive approach that simultaneously quantified 8-epiPGF2α, a biomarker of oxidative stress and biomarkers of cigarette smoke exposure, and incorporated these biomarkers into the statistical analysis, and demonstrated an additive effect of 8-epiPGF2α on lung cancer risk. Third, the prospective study design eliminated potential effects of lung cancer development and progression on 8-epiPGF2α. Fourth, the long term follow-up of study subjects after baseline urine collection allowed for a detailed analysis for the potential impact of subclinical stage and progression of disease (i.e. lung cancer) on urinary levels of 8-epiPGF2α and on the association between 8-epiPGF2α and lung cancer risk. Fifth, the study has the largest sample size so far with 610 lung cancer cases and 610 individually matched controls allowing for statistical analyses for the association between 8-epiPGF2α and lung cancer risk in never, former and current smokers, and among current smokers by different levels of cigarette smoking or smoking biomarkers. The major limitation is that the present study included men only.
In summary, using prospectively collected urine samples from participants in the Shanghai Cohort Study, we demonstrated that high urinary levels of 8-epiPGF2α at baseline were associated with a statistically significant, more than doubled risk of lung cancer in former and current smokers. The positive association remained after comprehensive adjustment for smoking intensity and duration, and multiple biomarkers of cigarette smoke exposure including total cotinine, total NNAL and PheT, suggesting an important effect of 8-epiPGF2α on the risk of lung cancer development, most likely due to tumor promotion, co-carcinogenesis and/or inflammation. Thus, smokers who are at a heightened state of oxidative stress in response to the insults of tobacco smoking may be more susceptible to smoking-induced lung carcinogenesis.
Supplementary material
Supplementary materials can be found at Carcinogenesis online.
Funding
USPHS grants R01CA 043092, R01 CA129534, R01 CA144034, R01 CA81301 and UM1 CA182876. Mass spectrometry was carried out in the Analytical Biochemistry Shared Resource of the Masonic Cancer Center, supported in part by Cancer Center Support Grant CA-77598.
Acknowledgements
We thank Xue-Li Wang of the Shanghai Cancer Institute for supervising the field work of the Shanghai Cohort Study. We also thank the Shanghai Cancer Registry for assistance with the identification of lung cancer cases via database linkage. We thank Eric Lake for assistance in the analysis of the urinary isoprostane.
Abbreviations
- BMI
body mass index
- CI
confidence interval
- 8-epiPGF2α
8-epi-prostaglandin F2α
- NNAL
4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol
- OR
odds ratio
- PheT
phenanthrene tetraol
References
- 1. Forman D., et al. (2014) The global and regional burden of cancer. In Stewart B.Wet al. (eds) World Cancer Report 2014. International Agency for Research on Cancer, Lyon, France, pp. 16–53. [Google Scholar]
- 2. Siegel R.L., et al. (2015) Cancer statistics, 2015. CA. Cancer J. Clin., 65, 5–29. [DOI] [PubMed] [Google Scholar]
- 3. International Agency for Research on Cancer(2004)Tobacco Smoke and Involuntary Smoking. IARC Monographs on the Evaluation of Carcinogenic Risk to Humans. IARC Scientific Publications, Lyon, France, vol. 83, pp. 36–1187. [PMC free article] [PubMed] [Google Scholar]
- 4. Jemal A., et al. (2011) Global cancer statistics. CA. Cancer J. Clin., 61, 69–90. [DOI] [PubMed] [Google Scholar]
- 5. Garcia M., et al. (2007)Global Cancer Facts & Figures 2007. American Cancer Society, Atlanta, GA, pp. 29–37. [Google Scholar]
- 6. Hecht S.S. (2012) Research opportunities related to establishing standards for tobacco products under the family smoking prevention and tobacco control act. Nicotine Tob. Res., 14, 18–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hecht S.S. (2003) Tobacco carcinogens, their biomarkers and tobacco-induced cancer. Nat. Rev. Cancer, 3, 733–744. [DOI] [PubMed] [Google Scholar]
- 8. International Agency for Research on Cancer(2007)Smokeless Tobacco and Tobacco-Specific Nitrosamines. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC Scientific Publications, Lyon, France, vol. 89, pp. 421–583. [Google Scholar]
- 9. Yuan J.M., et al. (2009) Urinary levels of tobacco-specific nitrosamine metabolites in relation to lung cancer development in two prospective cohorts of cigarette smokers. Cancer Res., 69, 2990–2995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Yuan J.M., et al. (2011) Urinary levels of cigarette smoke constituent metabolites are prospectively associated with lung cancer development in smokers. Cancer Res., 71, 6749–6757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wynder E.L., et al. (1967)Tobacco and Tobacco Smoke. Academic Press, New York, pp. 179–317. [Google Scholar]
- 12. Hoffmann D., et al. (1978) Tobacco carcinogenesis. In Gelboin H.Vet al. (eds) Polycyclic Hydrocarbons and Cancer. Academic Press, New York, pp. 85–117. [Google Scholar]
- 13. Wei H., et al. (1992) Suppression of tumor promoter-induced oxidative events and DNA damage in vivo by sarcophytol A: a possible mechanism of antipromotion. Cancer Res., 52, 2298–2303. [PubMed] [Google Scholar]
- 14. International Agency for Research on Cancer(2004)Tobacco Smoke and Involuntary Smoking. IARC Monographs on the Evaluation of Carcinogenic Risk to Humans. IARC Scientific Publications, Lyon, France, vol. 83, pp. 973–1003. [PMC free article] [PubMed] [Google Scholar]
- 15. Curtin G.M., et al. (2004) Short-term in vitro and in vivo analyses for assessing the tumor-promoting potentials of cigarette smoke condensates. Toxicol. Sci., 81, 14–25. [DOI] [PubMed] [Google Scholar]
- 16. Hecht S.S. (2005) Carcinogenicity studies of inhaled cigarette smoke in laboratory animals: old and new. Carcinogenesis, 26, 1488–1492. [DOI] [PubMed] [Google Scholar]
- 17. Takahashi H., et al. (2010) Tobacco smoke promotes lung tumorigenesis by triggering IKKbeta- and JNK1-dependent inflammation. Cancer Cell, 17, 89–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Milne G.L., et al. (2005) F2-isoprostanes as markers of oxidative stress in vivo: an overview. Biomarkers, 10 (Suppl 1), S10–S23. [DOI] [PubMed] [Google Scholar]
- 19. Il’yasova D., et al. (2010) Urinary biomarkers of oxidative status in a clinical model of oxidative assault. Cancer Epidemiol. Biomarkers Prev., 19, 1506–1510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Richelle M., et al. (1999) Urinary isoprostane excretion is not confounded by the lipid content of the diet. FEBS Lett., 459, 259–262. [DOI] [PubMed] [Google Scholar]
- 21. Liu J., et al. (2011) Relationship between biomarkers of cigarette smoke exposure and biomarkers of inflammation, oxidative stress, and platelet activation in adult cigarette smokers. Cancer Epidemiol. Biomarkers Prev., 20, 1760–1769. [DOI] [PubMed] [Google Scholar]
- 22. Dalaveris E., et al. (2009) VEGF, TNF-alpha and 8-isoprostane levels in exhaled breath condensate and serum of patients with lung cancer. Lung Cancer, 64, 219–225. [DOI] [PubMed] [Google Scholar]
- 23. Epplein M., et al. (2009) Association of plasma micronutrient levels and urinary isoprostane with risk of lung cancer: the multiethnic cohort study. Cancer Epidemiol. Biomarkers Prev., 18, 1962–1970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Yuan J.M., et al. (1996) Morbidity and mortality in relation to cigarette smoking in Shanghai, China. A prospective male cohort study. JAMA, 275, 1646–1650. [PubMed] [Google Scholar]
- 25. Yuan J.M., et al. (2001) Prediagnostic levels of serum beta-cryptoxanthin and retinol predict smoking-related lung cancer risk in Shanghai, China. Cancer Epidemiol. Biomarkers Prev., 10, 767–773. [PubMed] [Google Scholar]
- 26. Yan W., et al. (2007) Quantitation of isoprostane isomers in human urine from smokers and nonsmokers by LC-MS/MS. J. Lipid Res., 48, 1607–1617. [DOI] [PubMed] [Google Scholar]
- 27. London S.J., et al. (2000) Isothiocyanates, glutathione S-transferase M1 and T1 polymorphisms, and lung-cancer risk: a prospective study of men in Shanghai, China. Lancet, 356, 724–729. [DOI] [PubMed] [Google Scholar]
- 28. Yuan J.M., et al. (2014) Urinary metabolites of a polycyclic aromatic hydrocarbon and volatile organic compounds in relation to lung cancer development in lifelong never smokers in the Shanghai Cohort Study. Carcinogenesis, 35, 339–345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Yuan J.M., et al. (2016) Genetic determinants of cytochrome P450 2A6 activity and biomarkers of tobacco smoke exposure in relation to risk of lung cancer development in the Shanghai cohort study. Int. J. Cancer, 138, 2161–2171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Rothman K.J. (1976) The estimation of synergy or antagonism. Am. J. Epidemiol., 103, 506–511. [DOI] [PubMed] [Google Scholar]
- 31. Dai Q., et al. (2009) Oxidative stress, obesity, and breast cancer risk: results from the Shanghai Women’s Health Study. J. Clin. Oncol., 27, 2482–2488. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Wu H.C., et al. (2008) Urinary 15-F2t-isoprostane, aflatoxin B1 exposure and hepatitis B virus infection and hepatocellular carcinoma in Taiwan. Carcinogenesis, 29, 971–976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Barocas D.A., et al. (2011) Oxidative stress measured by urine F2-isoprostane level is associated with prostate cancer. J. Urol., 185, 2102–2107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Morrow J.D., et al. (1995) Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers. Smoking as a cause of oxidative damage. N. Engl. J. Med., 332, 1198–1203. [DOI] [PubMed] [Google Scholar]
- 35. Reilly M., et al. (1996) Modulation of oxidant stress in vivo in chronic cigarette smokers. Circulation, 94, 19–25. [DOI] [PubMed] [Google Scholar]
- 36. Pilz H., et al. (2000) Quitting cigarette smoking results in a fast improvement of in vivo oxidation injury (determined via plasma, serum and urinary isoprostane). Thromb. Res., 99, 209–221. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.