Fax: (011) 81-3-5685-3075
Susceptibility to lung cancer and genetic polymorphisms in the alcohol metabolite-related enzymes alcohol dehydrogenase 3, aldehyde dehydrogenase 2, and cytochrome P450 2E1 in the Japanese population
Article first published online: 8 JUN 2007
Copyright © 2007 American Cancer Society
Volume 110, Issue 2, pages 353–362, 15 July 2007
How to Cite
Minegishi, Y., Tsukino, H., Muto, M., Goto, K., Gemma, A., Tsugane, S., Kudoh, S., Nishiwaki, Y. and Esumi, H. (2007), Susceptibility to lung cancer and genetic polymorphisms in the alcohol metabolite-related enzymes alcohol dehydrogenase 3, aldehyde dehydrogenase 2, and cytochrome P450 2E1 in the Japanese population. Cancer, 110: 353–362. doi: 10.1002/cncr.22795
Fax: (011) 81-3-5685-3075
- Issue published online: 29 JUN 2007
- Article first published online: 8 JUN 2007
- Manuscript Accepted: 15 MAR 2007
- Manuscript Revised: 14 MAR 2007
- Manuscript Received: 2 JUL 2006
- Cancer Research from the Ministry of Health, Labor, and Welfare of Japan
- lung cancer;
- alcohol consumption;
- case-control study;
- genetic polymorphism;
- alcohol dehydrogenase 3;
- aldehyde dehydrogenase 2;
- cytochrome P450 2E1
It is believed that acetaldehyde plays an important role in alcohol-related carcinogenesis; although current epidemiologic studies have provided inconsistent findings on the association between alcohol consumption and the risk of lung cancer.
To clarify the hypothesis that genetic polymorphisms in alcohol-metabolizing enzymes may influence susceptibility to lung cancer, the authors conducted a hospital-based case-control study and examined genetic polymorphisms in the alcohol dehydrogenase 3, aldehyde dehydrogenase 2 (ALDH2), and cytochrome P450 2E1 genes in 505 patients with histologically confirmed lung cancer and in a group of 256 noncancer controls who provided complete cigarette and alcohol consumption histories. Genotyping was conducted by polymerase chain reaction-restriction fragment-length polymorphism assay.
A significant association was noted between alcohol consumption and lung cancer risk. Thus, using the median value for the controls as the cut-off point, the odds ratios (OR) for light and heavy drinkers were 1.76 and 1.95, respectively (P for trend = .012), compared with nondrinkers. In addition, there was a significant trend toward increased risk of lung cancer in drinkers with ALDH2 variant alleles (P for trend <.0001). The adjusted OR for heavy drinkers was 6.15 compared with nondrinkers. Regarding associations between histologic type and genotypes, the ALDH2 variant allele was significantly less common in patients who had adenocarcinoma compared with controls.
The current observations suggested a positive association between alcohol consumption and the risk of lung cancer: Drinking may increase the risk, especially among individuals who have the variant ALDH2 alleles. Cancer 2007. © 2007 American Cancer Society.
Epidemiologic studies have provided inconsistent results regarding the associations between alcohol consumption and the risk of lung cancer. In general, therefore, the involvement of alcohol in lung cancer etiology has been regarded with skepticism, with any indication of an association being attributed in most instances to confounding factors, such as cigarette smoking.1 It indeed is difficult to separate the effects of alcohol and smoking because, the 2 tend to be correlated, but this problem does not automatically exclude the possibility that there is a separate alcohol effect. A panel of experts commissioned by the World Cancer Research Fund and the American Institute for Cancer Research in 1997, after reviewing the epidemiologic evidence, concluded that alcohol intake possibly may increase lung cancer risk.2 Although the mechanism by which alcohol may cause cancer remains obscure, many epidemiologic studies have identified chronic alcohol consumption as a significant risk factor for cancers of the oral cavity, pharynx, larynx, and esophagus in humans.3 When investigating the role of alcohol-related carcinogenesis, most studies have concentrated on the type of alcoholic beverage consumed and the amount of daily intake, but this does not fully explain the variance in individual susceptibility to alcohol-related cancer.
Recent reports strongly implicate acetaldehyde, the first metabolite of ethanol, rather than alcohol itself, as responsible for the risk of developing alcohol-related cancers. It has been reported that acetaldehyde causes mutations by DNA adduct formation and inhibition of DNA repair. Moreover, drinking or inhaling acetaldehyde has mutagenic and carcinogenic effects and induced nasal and laryngeal carcinomas in experimental animals.4–8
Ethanol is primarily (80%) oxidized to acetaldehyde by alcohol dehydrogenase (ADH), and most of this acetaldehyde is then eliminated by aldehyde dehydrogenase (ALDH). However, ethanol and acetaldehyde also are metabolized through the microsomal ethanol-oxidizing system and the microsomal acetaldehyde-oxidizing system, and cytochrome P450 2E1 (CYP2E1) is a major contributor to those systems.9, 10CYP2E1 has high oxidation activity and is induced by long-term alcohol intake. These enzymes exhibit wide interindividual variability in their activity, suggesting that the variation may be caused by genetic polymorphisms.
There are several ADH subtypes, some of which have genetic variants with altered kinetic properties. ADH3 is polymorphic, and the enzyme encoded by the ADH allele metabolizes ethanol to acetaldehyde 2.5 times faster than that encoded by the ADH allele.11ALDH2 is a key enzyme in the elimination of acetaldehyde. In individuals with ALDH, a variant allele that is prevalent among East Asians (eg, ≈ 50% prevalence in Japan12), the activity of this enzyme is extremely low. The CYP2E1 variant allele, which is detectable by RsaI digestion (termed the c2 variant), corresponds to higher activity ethanol metabolism and is associated with greater alcohol consumption.13–15 Individuals who have 1 or more ADH, ALDH, and CYP2E1 c2 alleles accumulate more acetaldehyde in the blood after drinking ethanol and may be at increased risk for various alcohol-related diseases at similar levels of alcohol intake as individuals who do not carry these alleles. Because the ADH3 variant allele is common in whites, and the ALDH2 and CYP2E1 variant alleles are found at high frequency in Asians, research on these genes is most advanced regarding alcohol-related diseases and alcohol metabolism.
The association between genetic polymorphisms in these enzymes and susceptibility to some types of cancer has been reported in case-control studies. The ADH and ALDH alleles are associated closely with alcohol-related cancers in the upper aerodigestive tract,16–21 and systemic acetaldehydemia has been considered responsible for carcinogenesis in this locality. However, to our knowledge, there are no reports on associations between polymorphisms of ALDH and lung cancer risk. In relation to ADH, a negative association between genetic variation in ADH3 and lung cancer has been reported recently.22CYP2E1 is responsible primarily for the bioactivation of many low-molecular-weight, tobacco-specific carcinogens, including certain nitrosamines, such as N-nitrosodimethylamine and N-nitrosonornicotine. It is possible that the CYP2E1 c2 variant not only may increase the blood concentration of acetaldehyde but also may activate these carcinogens more strongly. Activated nitrosamines have been linked to the development of numerous cancers. However, results from studies that evaluated the role of CYP2E1 polymorphisms in relation to lung cancer have been discrepant.23–28 Because previous investigations did not adjust for alcohol consumption and/or did not have sufficient power to distinguish the risk from alcohol consumption, these inconsistent findings may have been caused by variations in CYP2E1 enzyme activity induced by ethanol.
We conducted a hospital-based case-control study to evaluate whether ADH3, ALDH2, or CYP2E1 polymorphisms are associated with lung carcinogenesis. The primary endpoint of the current study was to clarify the association between each genetic polymorphism and the risk of lung cancer, controlling for the amount of alcohol consumed and smoking habits. Furthermore, associations between alcohol consumption and lung cancer risk in individuals with variant alleles, again controlling for smoking, and associations between these polymorphisms and histologic characteristics were evaluated.
MATERIALS AND METHODS
This study was approved by the Institutional Review Board and the Ethics Committee of the National Cancer Center, Japan. The majority of eligible participants in this study were residents of Chiba and East Tokyo, and all were of Japanese nationality. Personal and clinical data from patients who participated in the Lung Cancer Database Project at the National Cancer Center Hospital East (NCCH-E) and the National Cancer Center Research Institute East were used in the current study. The database includes information on demographic factors, physical symptoms, psychological factors, and lifestyle factors (diet, smoking, etc) obtained from self-reported questionnaires and medical information from the patients' medical charts and blood, DNA, and urine specimens. All patients who were enrolled in the current study had primary lung cancer that was newly diagnosed with histologic or cytologic confirmation at the Thoracic Oncology Division of the NCCH-E, Japan, from September 1997 to June 2000. All patients provided their written informed consent prior to enrolment in this project. Unmatched controls were newly recruited individuals from the population with no history of cancer or other tumors who visited the Thoracic Oncology Division of NCCH-E from March 2002 to May 2003 and were confirmed as cancer-free by appropriate examinations (chest computed tomography scans, bronchofibroscopy, video-assisted thoracoscopic biopsy, etc). The major reasons for visiting the hospital were suspicions of lung cancer on chest x-ray or sputum cytology at their annual medical check-up or referral from other hospitals. Epidemiologic data were collected by personal interview. All individuals in the control group completed the same standardized questionnaire that was completed by the Lung Cancer Database Project participants, including detailed demographic information, history of cancer, occupational and residential history, and detailed information regarding alcohol and tobacco consumption. All participants provided their written consent.
Sample Collection and DNA Extraction
Four milliliters of peripheral venous blood were collected into heparinized tubes. Genomic DNA was purified from peripheral blood lymphocytes using a DNA isolation kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions and was stored at −80°C.
ADH3 and ALDH2 genotyping was performed by using the polymerase chain reaction-restriction fragment-length polymorphism (PCR-RFLP) method. To prevent the amplification of closely related ADH1 and ADH2 genes, samples initially were digested with the NlaIII restriction enzyme (TOYOBO, Osaka, Japan). A 145-base pair (bp) section of the ADH3 gene was amplified by PCR using 200 ng of predigested genomic DNA with primers (sense, 5′-GCTTTAAGAGT AAATATTCTGTCCCC-3′; antisense, 5′-AATCTACCTCT zTTCCGAAGC-3′). The PCR product obtained in this manner then was digested directly with restriction enzyme SspI (TOYOBO). After polyacrylamide gel electrophoresis, ADH3 alleles were visualized by ethidium bromide and were photographed under ultraviolet light. The ADH31 allele produced fragments of 67 bp, 63 bp, and 15 bp; and the ADH allele produced fragments of 131 bp and 15 bp.
A 134-bp fragment of the ALDH2 gene was amplified by PCR according to a slightly modified method of Harada et al.12 One hundred fifty nanograms of genomic DNA were mixed with 5 pmol of each primer (sense, 5′-CAAATTACAGGGTCAAGGGCT-3′; antisense: 5′-CCACACTCACAGTTTTCTCTT-3′) in a total volume of 50 μL that contained 50 μM deoxynucleotide triphosphate, 1.5 mM MgCl2, and 1 U Taq DNA polymerase; Takara Shuzo, Kyoto, Japan). Thirty-five cycles (denaturation at 94°C for 15 seconds, annealing at 58°C for 1 minute and 30 seconds, and polymerization at 72°C for 30 seconds) were performed using a GeneAmp PCR system 9600 (PerkinElmer, Oak Book, Ill). After purification, each PCR product was digested with MboII (TOYOBO), electrophoresed on a 20% polyacrylamide gel, stained with ethidium bromide, and photographed. The ALDH allele produced fragments of 125 bp and 9 bp, and the ALDH allele produced fragments of 134 bp.
The CYP2E1 genotypes ascribed to the RsaI site in the 5′-flanking region also were identified as RFLPs by PCR. Genomic DNA (100 ng) was subjected to PCR with each primer (sense, 5′-ATCCACAAGTG ATTTGGCTG-3′; antisense, 5′-CTTCATACAGACCCTC TTCC-3′). PCR was performed for 35 cycles under the following conditions: 1 minute at 95°C for denaturation, 1 minute at 55°C for primer annealing, and 1 minute at 72°C for primer extension. The 412-bp fragment was digested with RsaI (TOYOBO). The products that were yielded were fragments with 360 bp and 50 bp for c1/c1; 360 bp, 50 bp, and 410 bp for c1/c2; and 410 bp for c2/c2 detected by electrophoretic analysis in 5% polyacrylamide gels.
Patient characteristic (see Table 1) were compared with characteristic in the control group by using the Student t test or the chi-square test. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were obtained by unconditional logistic regression analysis. In our regression models, we adjusted ORs for potential confounding variables, including age, sex, smoking status (never, past, current) or amounts smoked (pack-years) and alcohol consumed (none, light, heavy). Because differences in the amount of alcohol consumed (ethanol, in gram per day) were very large, we divided those who drank into 3 categories: nondrinkers, light drinkers (≤31.6 g per day), and heavy drinkers (>31.6 g per day). The amount of tobacco smoke exposure was calculated as pack-years (usual amount per day/20 × overall duration [years] of use). Participants were considered current smokers if they smoked up to 1 year before the date of diagnosis in the case group or up to the date of the interview for the control group. The average amount of daily ethanol intake was calculated in grams. Calculation of this value was based on an average ethanol content of 4-volume% in beer, 15-volume% in Japanese sake (rice wine), 25-volume% in Japanese spirits (syochu), 12-volume% in wine, and 40-volume% in spirits. Drinking frequency was assessed as 5 categories: less than once a week, 1 or 2 days a week, 3 or 4 days a week, 5 or 6 days a week, and daily. Categorical variables were compared with the chi-square test. ORs and 95% CIs were calculated by using logistic regression analysis adjusting for age, sex, smoking, and drinking. The Mantel extension test was used to evaluate linear trends across categories of alcohol consumption that were divided into 4 categories by quartiles for control. Resulting P values <.05 (2-tailed) were considered statistically significant. All statistical analyses were performed using the SAS statistical software package (SAS Institute Inc., Cary, NC).
|Characteristic||No. (%)||P for difference|
|Cases (n = 505)||Controls (n = 256)|
|Mean age ± SD, y||64.8 ± 8.3||63.5 ± 10.2||.06*|
|Men||360 (71.3)||126 (49.2)|
|Women||145 (28.7)||130 (50.8)||<.0001†|
|Never||140 (27.7)||129 (50.4)|
|Past||97 (19.2)||64 (25)|
|Current||268 (53.1)||63 (24.6)||<.0001†|
|Smoking amounts, pack-years|
|<27||35 (36.1)||32 (50)|
|≥27||62 (63.9)||32 (50)||.08†|
|<40||71 (26.5)||30 (47.6)|
|≥40||197 (73.5)||33 (52.4)||.001†|
|Alcohol drinking habit, times/wk|
|Seldom||116 (23)||118 (46.1)|
|≤2||43 (8.5)||42 (16.4)|
|3–6||96 (19)||22 (8.6)|
|Daily||250 (49.5)||74 (28.9)||≤.0001†|
|Alcohol amounts, g/day|
|0||120 (23.8)||119 (46.5)|
|<31.6||154 (30.5)||65 (25.4)|
|≥31.6||231 (45.7)||72 (28.1)||≤.0001†|
In total 510 patients with lung cancer (cases) and 260 healthy controls participated in this study. Because of the lack of DNA samples or information on lifestyle, 9 participants were eliminated. Table 1 summarizes the baseline characteristics of selected variables for the lung cancer cases and controls. Age distribution was similar in both groups (mean, 64.8 years and 63.5 years, respectively); however, the cases were more likely than the controls to be men (71.3% and 49.2%), to be current smokers (53.1% and 24.6%) and heavy smokers, and to consume more alcohol. The proportions of those who consumed >31.6 g per day of ethanol and of daily drinkers were 45.7% and 49.5%, respectively, for cases and 28.1% and 28.9%, respectively, for controls. The median values from the control group for the 2 smoking amount categories were used as the cut-off values. The 3 categories of alcohol consumption were lifetime nondrinker, below the median intake, and above the median intake.
The frequency of ADH3, ALDH2, and CYP2E1 genotypes and ORs among lung cancer cases and controls are presented in Table 2. After adjustment for age, sex, smoking amount, and amount of alcohol consumed, the ORs for individuals with the ADH3, ALDH2, and CYP2E1 variant alleles, compared with individuals who were homozygous for the common allele, were 1.01, 0.73, and 0.93, respectively. Thus, there were no significant differences in the frequencies of any genotypes between cases and controls. The OR for carriers of the CYP2E1 c2/c2 genotype, compared with the c1/c1 genotype, was 4.66 (P < .05). This genotype is not in Hardy-Weinberg equilibrium in the control population, the observed frequency is most likely an underestimate, and the finding of an association with lung cancer is most likely a false-positive result.
|Cases (n = 505)||Controls (n = 256)||Crude||Adjusted*|
|C/C||459 (90.9)||227 (88.7)||1||1|
|C/V||44 (8.7)||29 (11.3)||0.75 (0.46–1.23)||0.71 (0.40–1.16)|
|V/V||2 (0.4)||0 (0)||—||—|
|C/V and V/V||46 (9.1)||29 (11.3)||0.78 (0.48–1.28)||0.74 (0.44–1.24)|
|C/C||319 (63.2)||134 (52.3)||1||1|
|C/V||168 (33.3)||108 (42.2)||0.65 (0.48–0.90)†||0.73 (0.52–1.03)|
|V/V||18 (3.6)||14 (5.5)||0.54 (0.26–1.12)||0.75 (0.35–1.59)|
|C/V and V/V||186 (36.8)||122 (47.7)||0.64 (0.47–0.87)†||0.73 (0.53–1.02)|
|C/C||300 (59.4)||147 (57.4)||1||1|
|C/V||175 (34.7)||106 (41.4)||0.81 (0.59–1.11)||0.83 (0.60–1.15)|
|V/V||30 (5.9)||3 (1.2)||4.90 (1.47–16.32)†||4.66 (1.36–16.0)†|
|C/V and V/V||205 (40.6)||109 (42.6)||0.92 (0.68–1.25)||0.93 (0.68–1.29)|
Without taking these genotypes into consideration, a direct association between alcohol consumption and lung cancer occurrence can be derived, as shown in Table 3. Drinking was classified as none, light (≤31.6 g per day) or heavy (>31.6 g per day). When adjusted for age, sex, and smoking amounts, drinking imposed a significantly greater risk of lung cancer occurrence. The ORs for the light drinkers and heavy drinkers, compared with nondrinkers, were 1.76 and 1.95, respectively (P for trend = .012). Thus, the risk of lung cancer increases as the amount alcohol consumed increases.
|Genotype||Drinkers||P for trend‡|
|Nondrinkers||≤≤31.6 g/Day||>31.6 g/Day|
|No.*||Reference||No.*||OR (95% CI)†||P||No.*||OR (95% CI)†||P|
|All||120/119||1||154/65||1.76 (1.12–2.75)||.014||231/72||1.95 (1.19–3.21)||.0085||.012|
|C/C||112/105||1||141/60||1.59 (0.99–2.55)||.054||206/62||1.88 (1.10–3.21)||.02||.025|
|C/V and V/V||8/14||1||13/5||4.31 (0.912–20.38)||.065||25/10||3.28 (0.742–14.55)||.12||.17|
|C/C||57/41||1||99/39||0.75 (0.39–1.42)||.37||163/54||0.46 (0.2– 0.99)||.049||.03|
|C/V and V/V||63/78||1||55/26||3.63 (1.76–7.46)||.0005||68/18||6.15 (2.77–13.65)||<.0001||<.0001|
|C/C||72/61||1||95/36||1.81 (0.97–3.38)||.061||133/50||1.67 (0.86–3.21)||.13||.31|
|C/V and V/V||48/58||1||59/29||1.74 (0.91–3.35)||.097||98/22||2.56 (1.16–5.65)||.02||.005|
ORs for developing lung cancer in association with the ADH3, ALDH2, and CYP2E1 genotypes also are presented in Table 3. Similar to what was observed in all participants taken together, an increased risk for developing lung cancer also was observed among individuals who were homozygous for the common allele ADH. However, because there were too few ADH3 variant allele carriers to analyze any association between alcohol consumption and lung cancer risk for this allele, it was inappropriate to compare the ADH and ADH genotypes.
The adjusted OR for the ALDH group was 0.75 (95% CI, 0.39–1.42) in light drinkers and 0.46 (95% CI, 0.20–0.99) in heavy drinkers. In contrast, individuals with the ALDH allele had a significantly greater risk of lung cancer; light drinkers had a 3.6-fold increased risk, and heavy drinkers had a 6.2-fold increased risk compared with nondrinkers (P for trend < .0001). These results indicate that, in individuals with the ALDH2 variant allele, continuous alcohol consumption is a strong risk factor for lung cancer.
The OR for the CYP2E1 c1/c1 genotype was 1.81 (95% CI, 0.97–3.38) for light drinkers and 1.67 (95% CI, 0.86–3.21) for heavy drinkers. For individuals with the CYP2E1 c2 allele, the OR was 1.74 (95% CI, 0.91–3.35) for light drinkers and 2.56 (95% CI, 1.16–5.65) for heavy drinkers (P for trend = .005). These results may indicate that individuals with the CYP2E1 variant allele are in a high-risk group for lung cancer in heavy drinkers.
It must be emphasized that, because of differences in distribution according to sex between cases and controls, we analyzed relative risks only in men (Table 4). For baseline characteristics among men, higher consumption of alcohol and more smoking were observed, as expected. Regarding associations between alcohol consumption and lung cancer risk, drinking was associated with an increased risk of developing lung cancer in all participants. The adjusted OR for the light and drinkers, compared with nondrinkers, was 6.54 (95% CI, 3.13–13.7) and 6.58 (95% CI, 3.28–13.2), respectively. However, in individuals with active ALDH genotypes, there was no association between alcohol consumption and lung cancer risk. In individuals with the inactive ALDH alleles, the risk for lung cancer was 6.8-fold (95% CI, 2.72–17.1) for light drinkers and 9.3-fold (95% CI, 3.72–23.4) for heavy drinkers compared with nondrinkers (P for trend <.0001). The risk in men who were heavy drinkers was much greater compared with women and those who carried the active ALDH21–1 genotype.
|Genotype||Drinkers||P for Trend‡|
|Nondrinkers||≤31.6 g/Day||>31.6 g/Day|
|No.*||Reference||No.*||OR (95% CI)†||P||No.*||OR (95% CI)†||P|
|All||17/31||1||120/36||6.54 (3.13–13.65)||<.0001||223/59||6.58 (3.28–13.22)||≤.0001||<.0001|
|C/C||15/27||1||110/34||6.14 (2.83–13.29)||<.0001||201/49||7.27 (3.44–15.36)||≤.0001||<.0001|
|C/V and V/V||2/4||1||10/2||23.31(1.41–286.0)||.028||22/10||5.43 (0.63–47.09)||.12||.47|
|C/C||5/2||1||72/16||1.47 (0.25–8.67)||.67||158/42||1.10 (0.20–6.23)||.91||.29|
|C/V and V/V||12/29||1||48/20||6.82 (2.72–17.13)||<.0001||65/17||9.33 (3.72–23.39)||≤.0001||<.0001|
|C/C||10/14||1||77/24||5.22 (1.95–13.94)||.0003||125/42||4.71 (1.85–12.05)||.0012||.08|
|C/V and V/V||7/17||1||43/12||8.31 (2.67–25.89)||.0001||98/17||9.93 (3.39–29.09)||≤.0001||<.0001|
In individuals with the c2 allele, the risk of lung cancer for light drinkers (OR, 8.31; 95% CI, 2.67–25.9) and for heavy drinkers (OR, 9.93; 95% CI, 3.39–29.1) was increased compared with individuals who were homozygous for the CYP2E1 c1 allele and compared with the risks in all men. However, it should be noted that, because of the low incidence of homozygosity for variant allele in the control group, statistical power was limited in this instance. Similar assessments also were made in women, but no significant associations between any genotype and lung cancer risk were observed (data not shown).
Table 5 shows the distribution of the ADH3, ALDH2, and CYP2E1 genotypes according to tumor histology. The frequency of the ADH allele for all histologic types was similar to the frequency observed in controls. The frequency of the ALDH allele for squamous cell carcinomas, small cell carcinomas, and other histologic types was similar to that observed in controls. However, the ALDH allele was significantly less common in patients with adenocarcinomas than in controls (36.1% vs 47.7%; P = .018). In contrast, the CYP2E1 c2/c2 genotype was more common in patients with adenocarcinomas (5.8%) and small cell carcinomas (9.8%) than in controls (1.2%).
|Control group (n = 256)||Histologic type|
|Adenocarcinoma (n = 330)||Squamous cell (n = 100)||Small cell (n = 51)||Other (n = 24)|
|C/C||227 (88.3)||297 (90)||91 (91)||48 (94.1)||23 (95.8)|
|C/V||29 (11.7)||31 (9.4)||9 (9)||3 (5.9)||1 (4.2)|
|V/V||0 (0)||2 (0.6)||0 (0)||0 (0)||0 (0)|
|P for difference*||.35||.52||.25||.28|
|C/C||134 (52.3)||211 (63.9)||54 (54)||36 (70.6)||18 (75)|
|C/V||108 (42.2)||104 (31.5)||45 (45)||13 (25.5)||6 (25)|
|V/V||14 (5.5)||15 (4.6)||1 (1)||2 (3.9)||0 (0)|
|P for difference*||.018||.17||.056||.083|
|C/C||147 (57.4)||197 (59.7)||59 (59)||31 (60.8)||13 (54.2)|
|C/V||106 (41.4)||114 (34.6)||37 (37)||15 (29.4)||9 (37.5)|
|V/V||3 (1.2)||19 (5.8)||4 (4)||5 (9.8)||2 (8.3)|
|P for difference*||.0067||.19||.001||.04|
In this study, we observed that alcohol consumption was an independent risk factor for lung cancer after adjusting for the influence of smoking (P for trend = .012). Although we assumed that individuals who had the ADH31–1 genotype were at greater risk for lung cancer compared with individuals who had the ADH allele, there was no evidence of an association between lung cancer and the ADH3 genotype in any analysis. Because the enzyme activity of ALDH is extremely low, acetaldehyde accumulates after alcohol intake. We could not demonstrate any association of ALDH2 genotypes with the risk of lung cancer after adjusting for smoking and the amount of alcohol consumed. However, we observed that individuals who had the ALDH allele were at a significantly greater risk of lung cancer because of alcohol consumption, although there was a significant trend for lower levels of alcohol consumption in individuals who had the ALDH genotype (P for trend = .03). We hypothesized that not only the differences in blood acetaldehyde concentrations but also the differences in enzyme activity on tobacco-specific carcinogens contribute to carcinogenesis. However, we produced no evidence that lung cancer risk is related to possession of the CYP2E1 c2/c2 genotype or that the CYP2E1 genotype modifies lung cancer susceptibility related to alcohol intake.
The control population for this study was recruited from the visitors to the NCCH-E. The majority of patients had false-positive chest x-rays at their annual check-up and had normal chest computed tomography scans, and they were not suffering from any respiratory illness. Furthermore, their family medical histories were similar to those expected in the ordinary Japanese population, although the number of current smokers among both men (42.9%) and women (6.9%) may have been somewhat lower than the average (46.8% and 11.1%, respectively, for 2003 according to the Announcement of the Ministry of Health, Labor, and Welfare). For these reasons, we believe that our control group was not at greater risk of cancer occurrence compared with the regular Japanese population. Moreover, it was not necessary to take into account any biases stemming from the selective inclusion only of consenting participants, because the great majority of both patients and controls agreed to participate in the study.
The data from the control group showed that individuals who had the ALDH2 wild-type genotype consumed more alcohol than individuals who had the variant genotype. This may suggest that genetic polymorphisms of alcohol-metabolizing enzymes influence drinking habits, because consumption may be limited by the unpleasant reactions caused by the accumulation of acetaldehyde in individuals with ALDH2 variant genotypes. Nonetheless, habitual drinking can increase consumption because of increased microsomal acetaldehyde-oxidizing system activation, further promoting the oxidation of acetaldehyde. The association between drinking habit and ADH3 and CYP2E1 genotypes remains uncertain.
Regarding correlations between smoking and drinking habits, the coexistence of smoking and drinking increased the risk of lung cancer compared with nondrinkers who never smoked, particularly the OR for heavy smokers (>37 pack-years) and drinkers, which was 8.4 (95% CI, 2.3–30.2; P = .0012) in the light drinkers and 7.0 (95% CI, 2.1–23.4) in the heavy drinkers (data not shown).
The involvement of alcohol in lung cancer etiology has been controversial, although many epidemiologic studies have suggested positive associations between different parameters of alcohol consumption and lung cancer risk. In the current study, we have demonstrated that drinking is a strong risk factor for lung cancer that is dose-dependent and is stronger in men than in women. This same tendency was observed even in the genotype analysis, but none of the results indicated a significant association between lung cancer and drinking in women. Furthermore, no associations were observed between peripheral lung adenocarcinoma, drinking, and genotypes of alcohol metabolite-related enzymes in women.
The question of ethnicity in the distribution of the polymorphisms of these alcohol metabolite-related enzyme genes always must be considered. The ADH allele is present in almost 60% of whites but is far more rare (5–10%) in Japanese. In contrast, the ALDH allele is found only in Asians. The CYP2E1 c2 allele is present in 35% to 56% of Japanese and Chinese, and in 2% to 5% of whites. In the current study, the frequency of variant alleles of each polymorphism was 9.9% for ADH3, 40.5% for ALDH2, and 41.3% for CYP2E1. This is consistent with previous studies in Japanese and other Asians.
We observed that the risk for lung cancer was increased significantly by alcohol consumption in a dose-dependent fashion in individuals with the ALDH alleles. Previously, some Japanese studies also showed a strong genetic and environmental interaction between ALDH and alcohol intake for the risk of developing esophageal and upper aerodigestive tract cancer.18–21 In contrast, for individuals with the ALDH genotype, there was an inverse association between alcohol consumption and the risk of lung cancer. These results suggest that increased acetaldehyde concentrations from a reduction in acetaldehyde oxidation caused by the presence of the ALDH allele contribute to the development of lung cancer. Significantly higher blood acetaldehyde concentrations after drinking in individuals with the ADH or ALDH allele have been reported compared with the concentrations in individuals who lacked these alleles,11, 29 and it has been demonstrated that breath acetaldehyde levels are proportional to blood acetaldehyde levels. Indeed, Muto et al.30 and Jones31 observed significantly higher acetaldehyde levels in the breath from individuals with the ALDH allele than in those without that allele. Therefore, exposure to higher concentrations of acetaldehyde in the lower respiratory tract may play a critical role in alcohol-related carcinogenesis. Regarding the influence of smoking, when adjusted for age, sex, and amount of alcohol consumed, the risks for developing lung cancer in current smokers were 1.5-fold greater for those with the inactive ALDH2 genotype (data not shown) compared with nonsmokers. The lung cancer risk for individuals with the ALDH allele was not increased further by smoking.
Although there have been some reports of a significant association between the ADH allele and some types of upper aerodigestive tract cancer, this association has been controversial.16, 17, 32–34 We failed to observe an association between ADH3 gene polymorphisms and the development of lung cancer, most likely because of the limited statistical power from the low frequency of the variant allele in the Japanese population.
Several investigations24, 31, 35, 36 have indicated that the CYP2E1 c2 allele is associated with susceptibility to some types of cancer. However, other investigators reported that carriers of the c2 allele had decreased susceptibility to a number of cancers25–27, 37 and reported no association between CYP2E1 genotypes and cancer.23, 28, 38 Discrepancies among these results may be caused by several factors, including differences in study design, sample size, and the populations' ethnicity. Statistical power usually is very limited in studies of the white population because of the extreme rarity of variant genotypes. Although CYP2E1 enzyme activity is induced by certain chemicals, such as ethanol, large interindividual variation has been observed in its constitutive activity as well as after induction. Watanabe et al.39 and Hayashi et al.15 reported that the RsaI variant c2 allele produced higher enzyme activity than the c1/c1 genotype in Japanese individuals, although this finding is itself controversial.40–42 Highly activated CYP2E1 induced by alcohol may play a more important role in the metabolic activation of several tobacco-specific procarcinogens, including various nitrosamines. It has been suggested that these low-molecular-weight carcinogens are associated with the development of peripheral adenocarcinoma. This finding is consistent with the results from our analysis of CYP2E1 presented in Table 5. However, the CYP2E1 c2/c2 genotype is not in Hardy-Weinberg equilibrium in the control population, the observed frequencies most likely are underestimates, and these findings of an association with histologic type most likely are false-positive results. In our analysis of ALDH2, the incidence of adenocarcinoma was high among individuals who had the wild-type genotype. Although a high incidence of squamous cell carcinoma was not observed, this result may imply that carcinogenesis caused by acetaldehyde occurs more in cancers other than adenocarcinoma as well as in esophageal and upper aerodigestive tract cancers.
A previous hospital-based study that was conducted in Japan failed to identify any association between the RsaI polymorphism and lung cancer, even when the analysis was stratified according to different histologic type.28 A more recent study indicated that there was a significant decrease in overall lung cancer risk associated with the possession of at least 1 copy of the CYP2E1RsaI variant allele, whereas there was no association between the CYP2E1RsaI polymorphism and the histologic type of lung cancer.27 However, none of the previous studies had adjusted for risk according to alcohol consumption levels, which strongly influence the activity of this enzyme. In the current study, we demonstrated that there is a difference between individuals who have the CYP2E1RsaI c2/c2 genotype compared with individuals who have the common c1/c1 genotype, with an adjusted OR of 4.66 (95% CI, 1.36–16.0) for the former group. Because of the low incidence of homozygosity in controls, the genotype distribution was not in Hardy-Weinberg equilibrium in our control population. The increased lung cancer risk among individuals with the CYP2E1 c2/c2 genotype likely was a false-positive result.
A correlation between the amount of alcohol consumed, genetic polymorphisms in the alcohol metabolite-related enzymes, and the stage of lung cancer was not observed in the current study, and we could not confirm that these factors were related to the aggressiveness of lung cancer. Furthermore, no associations were identified between the location of the primary cancer, the amount of alcohol consumed, and the genotype of these enzymes or between the risk for lung cancer and the type of alcoholic beverage consumed.
In summary, we report a significant association between amounts of alcohol consumed and susceptibility to lung cancer and that the risk of lung cancer in individuals with ALDH2 variant alleles, but not with ADH3 or CYP2E1 variant alleles, apparently was enhanced more by alcohol intake than in individuals with common genotypes. Moreover, to our knowledge, this is the first report documenting an association between lung cancer and genetic polymorphisms of alcohol metabolite-related enzymes. Because the sample size was relatively small for the investigation of effects stratified by each genotype, the current findings should be confirmed in large-scale studies with greater statistical power.
Supported in part by a Grant-in-Aid for Cancer Research from the Ministry of Health, Labor, and Welfare of Japan.
- 2Food, Nutrition and the Prevention of Cancer: A Global Perspective. American Institute for Cancer Research. Nutrition. 1999; 6: 523–526..
- 4International Agency for Research on Cancer. Allyl compounds, aldehyde, epoxies and peroxidies. IARC Monogr Eval Carcinog Risks Hum. 1985; 36: 101–132.
- 5A mutagenicity assessment of acetaldehyde. Mutat Res. 1998; 195: 1–20..
- 10Hepatic microsomal ethanol oxidizing system. J Biol Chem. 1970; 254: 2505–2512., .
- 27Association of CYP1A1, GSTM1, and CYP2E1 polymorphisms with lung cancer suggests cell type specificities to tobacco carcinogens. Cancer Res. 1998; 68: 4858–4863., , , et al.
- 29Genetic polymorphism of alcohol and aldehyde dehydrogenase and the effects on alcohol metabolism. Arukoru Kenkyuto Yakubutu Ison. 1993; 28: 3–25., , , .