Abstract
Background: Epidemiologic studies have shown the association between alcohol consumption and colorectal cancer, especially for rectal cancer. The alcohol related enzyme encoding gene ALDH2 has polymorphism Glu487Lys, and 487Lys allele is closely linked with phenotypic loss of enzyme activity. Materials and Methods: A hospital-based case-control study was conducted with 72 colon and 70 rectal cancer cases and 241 non-cancer controls to evaluate the alcohol consumption and ALDH2 Glu487Lys polymorphism. The logistic regression model was applied to estimate the odds ratios (ORs). Result: The crude ORs for Glu/Lys and Lys/Lys genotype relative to Glu/Glu for colon and rectal cancer were not statistically significant. However, with the rectal cancer analysis, the ORs for high alcohol consumption were greater with 487Glu/Lys genotype compared with Glu/Glu, albeit not. Conclusions: These observations suggested rectal cancer risk might be influenced by ALDH2 gene polymorphism. The prevention effect by alcohol reduction might differ by ALDH2 genotype.
Keywords: colorectal cancer, genetic predisposition of disease, alcohol, ALDH2 polymorphism
INTRODUCTION
Since the first pioneering study in England and Wales in 19571), many epidemiological investigations have been conducted to determine whether an association exists between alcohol and risk of colorectal cancer2,3). These studies suggest that alcohol consumption increases the risk, and a positive association has been more consistently observed for cancer of the rectum3).
Although several mechanisms can be hypothesized concerning alcohol effects, mucosal stimulation due to the metabolite, acetaldehyde, is regarded as of particular importance. Acetaldehyde bind with macromolecules and proteins thus forming acetaldehyde adducts4,5) which can interfere with DNA repair activity6) and can act as neoantigens with a subsequent immune response7). In the rat experiments, acetaldehyde causes hyperproliferation of rectal mucosal tissue8,9). A similar mechanism is also proposed for colonic mucosa, but the experimental evidence for this is lacking. Only one study showed the hyperregeneration of rectal mucosa in alcoholics, which indicating high exposure to alcohol and its metabolite may induce regeneration leading to carcinogenesis10). Although it needs further experimental studies, epidemiological evidence is consistent with the effect of acetaldehyde on rectal mucosa observed in rats.
Aldehyde dehydrogenase-2 (ALDH2) is a tetrameric enzyme which generates acetic acid from acetaldehyde and whose activity correlates with the in vivo concentrations of this derivative of alcohol11). A polymorphism of ALDH2 (Glu487Lys) is prevalent in Asians12), and the 487Lys allele is closely linked with phenotypic loss of enzyme activity. This polymorphism is located in a small three-stranded β-sheet domain that acts as an interface for tetramer formation and substitution with Lys at codon 487 affects the status of the ALDH2 enzyme13). Several of studies showed the correlation between phenotype such as drinking behavior I alcohol related symptoms and three genotypes by this polymorphism. The activity decreases in the order, Glu/Glu, Glu/Lys, and Lys/Lys14,15). Thus, there is a distinct possibility that the ALDH2 polymorphism could impact on the risk of colorectal cancer. Two studies in Japan have already examined this question, and both pointed to risk elevation in subjects having the ‘weak allele’16,17). Nevertheless, the gene-environment interaction between alcohol consumption and the ALDH2 genotype has not hitherto been detailed, to our knowledge. We therefore conducted a hospital based case-control study at Aichi Cancer Center (ACC).
METHODS
Study population and sample collection.
The case-control study was conducted as a part of series in a major project on genetic polymorphisms and cancer risk with patients at ACC18). Cases were recruited who were confirmed histologically to have colon (n=72) or rectal (n=70) cancer, excluding those with a history of other types of malignancies. Controls (n=241) were outpatients without any history of cancer who visited ACC during March to December 1999 for gastroscopy19); including 97 (40.2% out of 241) participants stated to be under medication for 107 diseases (not confirmed by their medical records); 23 with gastric/duodenal ulcer, another 23 for so-called gastritis, 16 with hypertension, 8 for pain including arthritis and lumbago, 7 with diabetes mellitus, 7 with hyperlipidemia, and other 23 miscellaneous diseases. All cases and controls were Japanese. The subjects who provided written informed consent for participation in this study were asked to complete a self-administered questionnaire and to provide blood from a peripheral vein. This study was approved by the Ethical Committee of ACC (Approval No. 12-23 and 12-27).
Environmental factors
Alcohol drinking was divided into three categories based on mainly drinking quantity; low (less than once a week), moderate (once a week or more frequently with less than 50mL of ethanol) or high (once a week or more frequently with 50mL of ethanol or more). Smokers were also divided into three categories (never, former, and current). We defined former smokers as those who had quit smoking more than 2 years before disease onset or the questionnaire study. For dietary factors, we asked frequency for four food items (whole meat, fish, raw vegetable, and tofu), three beverages (Japanese tea, black tea, and coffee), and preference for salt. We asked the cases to provide information about their life-style before the onset of disease, and the controls at the study enrollment.
Genotype analyses of the ALDH2
DNA of each subject was extracted from the buffy coat fraction with a QIAamp DNA Blood Mini Kit (Qiagen Inc., Valencia, CA). ALDH2 1543 G to A (accession no. NM_000690, Glu487Lys) was genotyped using PCR-CTPP (polymerase chain reaction with confronting two-pair primers) method developed in our laboratory20), as previously described21).
Statistical analysis
Accordance with the Hardy-Weinberg equilibrium (HWE), which indicates an absence of discrepancies between genotype and allele frequencies, was checked for controls with a χ2 test. Categorical variables were also tested in the same way. All odds ratios (ORs) and 95% confidence intervals (CIs) were estimated by an unconditional logistic regression model for each site of cancer and their combination. Age was adjusted as a continuous variable. All of the statistical analyses in this study were performed using STATA version 7.0 statistical software (STATA Corporation Inc., College Station, TX). Adjustment for multiple comparison was not performed, because the analyses were conducted in an exploratory context22), which requires a careful interpretation of any p-values.
RESULTS
A total of 142 cases and 241 controls were recruited. Their characteristics are shown in Table 1. Cases comprise 72 colon cancer and 70 rectal cancer patients. The age distribution was slightly older among colon cancer patients. The percentage of current male smokers was slightly higher in those with rectal cancer, but the distribution of smoking status was not statistically significant by χ2 test. High alcohol consumers were predominant in male rectal cancer cases with statistical significance (χ2 = 41.1, p <0.001).
Table 1. Background characteristics of cases and controls by sex.
| Males | Females | |||||||||||
|
|
|
|||||||||||
| Controls | Cases | Controls | Cases | |||||||||
|
|
|
|||||||||||
| Colon | Rectum | Colon | Rectum | |||||||||
| N=118 | N=42 | N=41 | N=123 | N=30 | N=29 | |||||||
| Number (%) of | ||||||||||||
| Age in years | ||||||||||||
| <40 | 2 | (1.7) | 3 | (7.1) | 0 | (0) | 0 | (0) | 2 | (6.7) | 2 | (6.9) |
| 40-49 | 21 | (17.8) | 7 | (16.7) | 6 | (14.6) | 23 | (18.7) | 2 | (6.7) | 5 | (17.2) |
| 50-59 | 34 | (28.8) | 12 | (28.6) | 16 | (39.0) | 56 | (45.5) | 12 | (40.0) | 9 | (31.0) |
| 60-69 | 61 | (51.7) | 12 | (28.6) | 15 | (36.6) | 44 | (35.8) | 8 | (26.7) | 10 | (34.5) |
| >70 | 0 | (0) | 8 | (19.1) | 4 | (9.8) | 0 | (0) | 6 | (20.0) | 3 | (10.3) |
| Years from diagnosis | ||||||||||||
| ≤ 3 year | - | 26 | (61.9) | 30 | (73.2) | - | 21 | (70.0) | 24 | (82.8) | ||
| > 3 year | - | 16 | (38.1) | 11 | (26.8) | 9 | (30.0) | 5 | (17.2) | |||
| Smoking status | ||||||||||||
| Never | 34 | (28.8) | 14 | (33.3) | 4 | (9.8) | 106 | (86.2) | 27 | (90.0) | 26 | (89.7) |
| Former | 38 | (32.2) | 18 | (42.9) | 18 | (43.9) | 5 | (4.1) | 1 | (3.3) | 0 | (0.0) |
| Current | 46 | (39.0) | 10 | (23.8) | 19 | (46.3) | 12 | (9.8) | 2 | (6.7) | 3 | (10.3) |
| Drinking status | ||||||||||||
| Low (< 1 time /week) | 35 | (29.7) | 9 | (21.4) | 6 | (14.6) | 101 | (82.1) | 26 | (86.7) | 23 | (79.3) |
| Moderate*1 | 50 | (42.3) | 17 | (40.5) | 12 | (29.3) | 17 | (13.8) | 3 | (10.0) | 5 | (17.2) |
| 1-4 times /week | 22 | 8 | 3 | 12 | 1 | 3 | ||||||
| ≥ 5 times /week | 28 | 9 | 9 | 5 | 2 | 2 | ||||||
| High*1 | 33 | (28.0) | 16 | (38.1) | 23 | (56.1) | 4 | (3.3) | 1 | (3.3) | 1 | (3.5) |
| 1-4 times /week | 8 | 3 | 1 | 3 | 0 | 1 | ||||||
| ≥ 5 times /week | 25 | 13 | 22 | 1 | 1 | 0 | ||||||
| Unknown | 0 | (0) | 0 | (0) | 0 | (0) | 1 | (0.8) | 0 | (0) | 0 | (0) |
*1 ‘Moderate’ indicates once per week or more frequently with less than 50 mL of ethanol and ‘High’ indicates once per week or more frequently with more than 50 mL.
The frequency of ALDH2 Glu/Glu, Glu/Lys and Lys/Lys genotypes were 52.3%, 39.8% and 7.9%, respectively for controls in total. The χ2 test for Hardy-Weinberg equilibrium with the polymorphism was not statistically significant (χ2 =0.014, p=0.905), indicating that no selective mechanisms for a specific genotype existed among the controls. When subjects were divided by sex, genotype distributions for total subjects or each site were not statistically different between males and females. The tests for HWE were not statistically significant for each sex, either (Table 2). The genotype frequencies for cases in total were 66.7%, 30.5%, and 2.8% for colon cancers, and 47.1%, 45.7%, and 5.7% for rectal cancers, the distribution among each site not differing significantly from that of controls. The ORs for ALDH2 Glu/Lys and Lys/Lys genotypes were not statistically significant. Point estimates for ALDH2 ORs showed a tendency for risk reduction with the 487Lys allele for colon (crude ORs compared with Glu/Glu: Glu/Lys 0.48 and Lys/Lys 0.23 for males; Glu/Lys 0.83 and Lys/Lys 0.36 for females), but not for rectal cancer (crude OR: Glu/Lys 1.07 and Lys/Lys 0.66 for males; Glu/Lys 1.71 and Lys/Lys 1.11 for females).
Table 2. Number of Cases and Controls, Odds Ratios (OR) and 95% CIs, for ALDH2 polymorphisms.
| Genotype | Males | Female | ||||||||||
|
|
|
|||||||||||
| Controls | Cases | Model*2 1 | Model 2 | Controls | Cases | Model 1 | Model 2 | |||||
|
|
|
|
|
|
|
|
|
|||||
| OR | 95%CI | OR | 95%CI | OR | 95%>CI | OR | 95%CI | |||||
| All cases | N=118 | N=83 | N=123 | N=59 | ||||||||
| Glu/Glu | 65(55.1%) | 53(63.9%) | 1.00 | Reference | 1.00 | Reference | 61(49.6%) | 28(47.5%) | 1.00 | Reference | 1.00 | Reference |
| Glu/Lys | 44(37.3%) | 26(31.3%) | 0.72 | 0.40-1.33 | 0.70 | 0.38-1.30 | 52(42.3%) | 28(47.5%) | 1.17 | 0.62-2.23 | 1.11 | 0.58-2.14 |
| Lys/Lys | 9(7.6%) | 3(3.6%) | 0.41 | 0.11-1.59 | 0.38 | 0.10-1.51 | 10(8.1%) | 3(5.1%) | 0.65 | 0.17-2.56 | 0.63 | 0.16-2.48 |
| UK*1 | 0(0.0%) | 1(1.2%) | - | 0(0.0%) | 0(0.0%) | |||||||
| Colon cancer | N=118 | N=42 | N=123 | N=30 | ||||||||
| Glu/Glu | 65(55.1%) | 31(73.8%) | 1.00 | Reference | 1.00 | Reference | 61(49.6%) | 17(56.7%) | 1.00 | Reference | 1.00 | Reference |
| Glu/Lys | 44(37.3%) | 10(23.8%) | 0.48 | 0.21-1.07 | 0.47 | 0.21-1.06 | 52(42.3%) | 12(40.0%) | 0.83 | 0.36-1.89 | 0.78 | 0.33-1.83 |
| Lys/Lys | 9(7.6%) | 1(2.4%) | 0.23 | 0.03-1.92 | 0.22 | 0.03-1.85 | 10(8.1%) | 1(3.3%) | 0.36 | 0.04-3.00 | 0.38 | 0.05-3.30 |
| Rectal cancer | N=118 | N=41 | N=123 | N=29 | ||||||||
| Glu/Glu | 65(55.1%) | 22(53.7%) | 1.00 | Reference | 1.00 | Reference | 61(49.6%) | 11(37.9%) | 1.00 | Reference | 1.00 | Reference |
| Glu/Lys | 44(37.3%) | 16(39.0%) | 1.07 | 0.51-2.27 | 1.11 | 0.51-2.38 | 52(42.3%) | 16(55.2%) | 1.71 | 0.73-4.00 | 1.61 | 0.26-4.31 |
| Lys/Lys | 9(7.6%) | 2(4.9%) | 0.66 | 0.13-2.27 | 0.65 | 0.12-3.35 | 10(8.1%) | 2(6.9%) | 1.11 | 0.21-5.77 | 1.03 | 0.20-5.39 |
| UK | 0(0.0%) | 1(2.4%) | - | 0(0.0%) | 0(0.0%) | |||||||
*1 UK indicates ALDH2 genotype unknown because DNA was not amplified by PCR. This subject was excluded from OR analyses.
*2 Model 1: crude OR. Model 2: adjusted for age and smoking status.
Alcohol consumption was not adjusted because of strong confounding with ALDH2 genotype.
Table 3 shows the ORs for the alcohol consumption according to ALDH2 genotype. Among individuals with ALDH2 Glu/Glu genotype, high alcohol consumption was linked with a high OR for rectal cancer (OR 3.41, 95% CI 0.92-12.6). The OR for high alcohol consumption was also higher than unity for colon cancer, but to a lesser extent. Similarly, a significantly high OR was observed among rectal cancer with Glu/Lys genotype (OR 8.07, 1.88-34.7), whereas no significant OR for colon cancer. Although ORs lower than unity were obtained for the moderate or high alcohol consumption among colon cancer cases, the numbers were limited in these groups compared with rectal cancer. Focusing on ALDH2 Lys/Lys type, there were no high alcohol consumption subjects. Although the OR for moderate alcohol consumption was extremely high for rectal cancers with Lys/Lys genotype, the number of cases in this group was limited. When focusing on males, similar trends were observed that the OR for high alcohol consumption among rectal cancer. The OR for males highly consuming alcohol with Glu/Glu genotype was 4.00 (0.45-35.5), whereas that with those with Glu/Lys genotype was 13.2 (2.24-78.0). On the other hand, analyses for females showed unstable estimation that might be due to small number of alcohol consumers among females.
Table 3. Adjusted*1 ORs and 95% CIs for alcohol drinking according to ALDH2 genotype.
| Alcohol consumption |
Controls | All cases | Colon cancer | Rectal cancer | ||||||
|
|
|
|
||||||||
| OR | 95%CI | OR | 95%CI | OR | 95%CI | |||||
| Total | N=240*2 | N=141*3 | N=72 | N=69*3 | ||||||
| ALDH2 487 Glu/Glu | ||||||||||
| N=126 | N=81 | N=48 | N=33 | |||||||
| Low | 54(42.9%) | 25(30.9%) | 1.00 | Reference | 17(35.4%) | 1.00 | Reference | 8(24.2%) | 1.00 | Reference |
| Moderate | 40(31.7%) | 24(29.6%) | 1.15 | 0.51-2.63 | 15(31.3%) | 0.94 | 0.35-2.47 | 9(27.3%) | 1.53 | 0.46-5.13 |
| High | 32(25.4%) | 32(39.5%) | 1.92 | 0.77-4.80 | 16(33.3%) | 1.21 | 0.41-3.52 | 16(48.5%) | 3.41 | 0.92-12.6 |
| P for trend: | 0.138 | 0.693 | 0.050 | |||||||
| ALDH2 487 Glu/Lys | ||||||||||
| N=95 | N=54 | N=22 | N=32 | |||||||
| Low | 64(67.4%) | 35(64.8%) | 1.00 | Reference | 17(77.3%) | 1.00 | Reference | 18(56.3%) | 1.00 | Reference |
| Moderate | 26(27.4%) | 10(18.5%) | 0.76 | 0.30-1.98 | 4(18.2%) | 0.53 | 0.14-1.97 | 6(18.8%) | 1.08 | 0.33-3.54 |
| High | 5(5.3%) | 9(16.7%) | 3.59 | 0.99-13.0 | 1(4.6%) | 0.69 | 0.07-6.89 | 8(25.0%) | 8.07 | 1.88-34.7 |
| P for trend: | 0.155 | 0.436 | 0.013 | |||||||
| ALDH2 487 Lys/Lys | ||||||||||
| N=19 | N=6 | N=2 | N=4 | |||||||
| Low | 18(94.8%) | 4(66.7%) | 1.00 | Reference | 1(50.0%) | 1.00 | Reference | 3(75.0%) | 1.00 | Reference |
| Moderate | 1(5.2%) | 2(33.3%) | 24.5 | 0.77-786.7 | 1(50.0%) | 18.2*4 | 0.58-569.7 | 1(25.0%) | 7.44 | 0.22-256.1 |
| High | 0(0.0%) | 0(0.0%) | NE*3 | 0(0.0%) | NE*5 | 0(0.0%) | NE*5 | |||
| P for trend: | 0.070 | NE*4 | 0.744 | |||||||
| Male | N=118 | N=82*3 | N=42 | N=40*3 | ||||||
| ALDH2 487 Glu/Glu | ||||||||||
| N=65 | N=53 | N=31 | N=22 | |||||||
| Low | 8(12.3%) | 5(9.4%) | 1.00 | Reference | 4(12.9%) | 1.00 | Reference | 1 (4.5%) | 1.00 | Reference |
| Moderate | 28(43.1%) | 18(34.0%) | 1.04 | 0.29-3.70 | 12(38.7%) | 0.88 | 0.22-3.49 | 6(27.3%) | 1.68 | 0.18-16.2 |
| High | 29(44.6%) | 30(56.6%) | 1.69 | 0.49-5.84 | 15(48.4%) | 1.07 | 0.27-4.17 | 15(68.2%) | 4.00 | 0.45-35.5 |
| P for trend: | 0.230 | 0.795 | 0.072 | |||||||
| ALDH2 487 Glu/Lys | ||||||||||
| N=44 | N=26 | N=10 | N=16 | |||||||
| Low | 19(43.2%) | 9(34.6%) | 1.00 | Reference | 5(50.0%) | 1.00 | Reference | 8(50.0%) | 1.00 | Reference |
| Moderate | 21(47.7%) | 8(30.8%) | 0.84 | 0.27-2.68 | 4(40.0%) | 0.74 | 0.17-3.19 | 4(25.0%) | 0.99 | 0.33-3.54 |
| High | 4(9.1%) | 9(34.6%) | 5.07 | 1.19-21.6 | 1(10.0%) | 0.97 | 0.09-10.9 | 4(25.0%) | 13.2 | 2.24-78.0 |
| P for trend: | 0.057 | 0.817 | 0.004 | |||||||
| ALDH2 487 Lys/Lys | ||||||||||
| N=9 | N=3 | N=1 | N=2 | |||||||
| Low | 8(88.9%) | 1(66.7%) | 1.00 | Reference | 0(0.0%) | 1.00 | Reference | 1(50.0%) | 1.00 | Reference |
| Moderate | 1(11.1%) | 2(33.3%) | 16.0 | 0.65-390.9 | 1(100.0%) | NE*5 | 1(50.0%) | 7.39 | 0.22-248.5 | |
| High | 0(0.0%) | 0(0.0%) | NE*4 | 0(0.0%) | NE*5 | 0(0.0%) | NE*5 | |||
| P for trend: | 0.090 | NE*5 | 0.265 | |||||||
| Female | N=123*2 | N=59 | N=30 | N=29 | ||||||
| ALDH2 487 Glu/Glu | ||||||||||
| N=61 | N=28 | N=17 | N=11 | |||||||
| Low | 46(75.4%) | 20(71.4%) | 1.00 | Reference | 13(76.5%) | 1.00 | Reference | 7 (63.6%) | 1.00 | Reference |
| Moderate | 12(19.7%) | 6(21.4%) | 1.32 | 0.42-4.15 | 3(17.7%) | 0.97 | 0.22-4.34 | 3(27.3%) | 1.69 | 0.37-7.75 |
| High | 3(4.9%) | 2(7.1%) | 2.75 | 0.37-20.8 | 1(5.9%) | 2.30 | 0.18-29.1 | 1(9.1%) | 2.44 | 0.18-32.8 |
| P for trend: | 0.331 | 0.679 | 0.390 | |||||||
| ALDH2 487 Glu/Lys | ||||||||||
| N=51 | N=28 | N=12 | N=16 | |||||||
| Low | 45(88.2%) | 26(92.9%) | 1.00 | Reference | 12(100.0%) | 1.00 | Reference | 14(87.5%) | 1.00 | Reference |
| Moderate | 5(9.8%) | 2(7.1%) | 0.69 | 0.12-3.91 | 0(0.0%) | NE*5 | 2(12.5%) | 1.26 | 0.20-7.58 | |
| High | 1(2.0%) | 0(0.0%) | NE*5 | 0(0.0%) | NE*5 | 0(0.0%) | NE*5 | |||
| P for trend: | 0.445 | NE*5 | 0.891 | |||||||
| ALDH2 487 Lys/Lys | ||||||||||
| N=10 | N=3 | N=1 | N=2 | |||||||
| Low | 10(100.0%) | 3(100.0%) | 1.00 | Reference | 1(100.0%) | 1.00 | Reference | 2(100.0%) | 1.00 | Reference |
| Moderate | 0(0.0%) | 0(33.3%) | NE*5 | 0(0.0%) | NE*5 | 0.58-569.7 | 0(0.0%) | NE*5 | ||
| High | 0(0.0%) | 0(0.0%) | NE*5 | 0(0.0%) | NE*5 | 0(0.0%) | NE*5 | |||
| P for trend: | NE*5 | NE*5 | NE*5 | |||||||
*1 For overall analyses, sex and age were adjusted. Age was adjusted in analyses for each sex.
*2 One controls was excluded from analysis because alcohol consumption data was not completed.
*3 One case was excluded from analysis because ALDH2 genotype was not determined.
*4 Adjusted OR for only age because subjects’ sex for Lys/Lys genotype and moderate drinking was male.
*5 NE indicates not estimated because case or controls were absent in these categories.
To evaluate the prognostic effect of the ALDH2 genotype among cases, we estimated the OR by the interval from diagnosis (incident group: ≤ 3 years; prevalent group; > 3 years). For the colorectal cancer, the ORs for having 487Lys allele were 0.79 (0.49-1.27) and 0.93 (0.47-1.84) for the incident and prevalent groups, respectively. Similar analyses according to subsite also showed no difference between two groups.
DISCUSSION
The present hospital based case-control study in Japan was conducted to evaluate the gene-environment interaction between alcohol and ALDH2 polymorphism for the risk of colorectal cancer. We observed that cancer of the rectum was more influenced by alcohol consumption than colon cancer, with increased risk among individuals with a ‘weak’ ALDH2 (Glu/Lys) genotype than those with the strongest (Glu/Glu) in terms of enzyme activity. This trend was obvious among men but not clear among women. The finding may be important with reference to differences in susceptibility to environment exposure among male individuals with different genetic traits.
Stronger harmful effects of alcohol drinking for ALDH2 487Lss allele carriers have been suggested for other cancers. Among Japanese alcoholics, the proportion of the 487Lys allele carriers was reported to be higher for those with cancer than for those without cancer, especially for patients with oropharyngolaryngeal and esophageal cancers16). We also found that a statistically significant interaction between heavy alcohol drinking and this ALDH2 polymorphism21). When the same amount of alcohol is drunk, the individuals with Glu/Lys genotype show a higher acetaldehyde concentration in serum and saliva than those with Glu/Glu genotype23,24). Accordingly, it is very likely in biological sense that the harmful effect would be stronger in 487Lys carriers drinking the same amount of alcohol.
To date, many epidemiological studies have evaluated the association between risk of colorectal cancers and alcohol consumption2,3). Risk elevation being more consistently observed for rectum. However, for the Japanese, the difference in the OR of alcohol consumption between colon and rectal cancer has been controversial25,26). Variation by subsite is supported by an animal experiment which demonstrated effects of alcohol only on rectum mucosa8).
We observed an approximately 3 times higher OR for alcohol consumption among the male rectal cancer subjects with Glu/Lys compared to those with Glu/Glu. Although the estimated interaction (2.27) among male was not statistically significant, the positive value suggested that Glu/Lys genotype might enhance the effect of alcohol in high consumer males. To our knowledge, this is the first documentation of a possible existence of interaction and from the viewpoint of prevention, the results have important implication. The number of drinkers among females was very small in our subjects (Table 1, 3), the observed interaction between alcohol and ALDH2 polymorphism may be interpretable only for male.
To date, two studies have concentrated attention on possible links between alcohol consumption and ALDH2 polymorphism for the risk of colorectal cancer16,17). Yokoyama et al evaluated the ALDH2 among alcoholics (46 colon cancer patients and 487 non-cancer controls) and reported OR of 3.4 (1.5-7.5) for the heterozygotes (Glu/Lys compared with Glu/Glu)16). Murata et al examined 265 colon cancer, 164 rectal cancer, and 121 non-cancer patients in the hospital-based prevalent case-control study. ORs of 2.13 (1.0-4.7) and 1.03 (0.5-2.2) were observed for ALDH2 487Lys allele carriers among colon and rectal cancers, respectively. Contrary to our result, the ORs for alcohol consuming were 3.1 (0.7-14.0) for colon cancer and 1.3 (0.2-7.0) for rectal cancer17). The reason for the inconsistency is unclear, but random differences in alcohol consumption and ALDH2 genotype distribution among controls might have affected the risk estimation.
We must say that there were several limitations to interpret this study results. Firstly, an attention needs to be paid to a prevalent case-control design. If the ALDH2 genotype under study has a prognostic effect, the ORs derived from prevalent case-control design would be influenced27). However, our result showed similar estimations for the ALDH2 genotype, suggesting that the estimated ORs for ALDH2 genotype and the interaction term were almost free from the prognostic effect. Secondly, the sample size for this study was not large enough to evaluate the interaction between alcohol consumption and ALDH2 genotype, thus the results obtained from this study are not conclusive. Thirdly, in our analyses the dietary habits such as fiber or fat intake whose relations between colorectal cancer were suggested could not be analyzed. The interaction between these habits and alcohol consumption, it was not adjusted in this study, may affect the observed result.
In conclusion, the present study provided evidence for the possible existence of a gene-environment interaction between alcohol consumption and ALDH2 genotype for the rectal cancer in Japanese. The prevention effect of alcohol reduction might be influenced by genotype of ALDH2. Although the interaction seems biologically plausible and analogous finding have been obtained for other sites of cancer, additional confirmatory studies are required with datasets for various ethnic groups.
ACKNOWLEDGMENT
The authors are grateful to Ms. Toshiko Saito, Ms. Naomi Takeuchi, Ms. Michiyo Tani, Ms. Keiko Asai and Ms. Hiroko Fujikura for their technical assistance. This work was supported in part by the Grant-in-Aid for Scientific Research on Priority Area (C) in 2000-2003 from the Ministry of Education, Science, Sports, Culture and Technology and the Grant-in-Aid for JSPS fellows. Keitaro Matsuo is supported by Research Fellowships of the Japan Society for the Promotion of Science for Young Scientist.
REFERENCES
- 1.Stocks P: Cancer incidence in North Wales and Liverpool region in relation to habits and environment. Brit Empire Cancer Campaign, 35th Annual Report, London, Suppl to Part 2, 1957: 1-127.
- 2.Kune GA, Vitetta L. Alcohol consumption and the etiology of colorectal cancer: a review of the scientific evidence from 1957 to 1991. Nutr Cancer, 1992; 18: 97-111. [DOI] [PubMed] [Google Scholar]
- 3.Potter JD. Nutrition and colorectal cancer. Cancer Causes Control, 1996; 7: 127-146. [DOI] [PubMed] [Google Scholar]
- 4.Barry RE, Williams AJ, McGivan JD. The detection of acetaldehyde/liver plasma membrane protein adduct formed in vivo by alcohol feeding. Liver, 1987; 7: 364-368. [DOI] [PubMed] [Google Scholar]
- 5.Lin RC, Smith RS, Lumeng L. Detection of protein-acetaldehyde adduct in the liver of rats fed alcohol chronically. J Clin Invest, 1988; 81: 615-619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Espina N, Lima V, Liever CS, Garro AJ. In vitro and in vivo inhibitory effect of ethanol and acetaldehyde on O6-methylguanine transferase. Carcinogenesis, 1988; 9: 761-766. [DOI] [PubMed] [Google Scholar]
- 7.Niemela O, Klajner F, Orrego H, et al. Antibodies against acetaldehyde modified protein epitopes in human alcoholics. Hepatology, 1987; 7: 1210-1214. [DOI] [PubMed] [Google Scholar]
- 8.Seitz HK, Simanowski UA, Garzon FT, et al. Possible role of acetaldehyde in ethanol-related rectal cocarcinogenesis in the rat. Gastroenterology, 1990; 98: 406-413. [DOI] [PubMed] [Google Scholar]
- 9.Simanowski UA, Suter P, Russell RM, et al. Enhancement of ethanol induced rectal mucosal hyper regeneration with age in F344 rats. Gut, 1994; 35: 1102-1106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Simanowski UA, Arce L, Knühl M, et al. Chronic alcohol consumption enhances rectal cell regeneration in man. Gastroenterology, 1994; 106; A442. (abstract) [Google Scholar]
- 11.Bosron WF, Li TK. Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases and their relationship to alcohol metabolism and alcoholism. Hepatology, 1986; 6: 502-510. [DOI] [PubMed] [Google Scholar]
- 12.Yoshida A, Huang IY, Ikawa M. Molecular abnormality of an inactive aldehyde dehydrogenase variant commonly found in Orientals. Proc Nat Acad Sci U S A, 1984; 81: 258-261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Steinmetz CG, Xie P, Weiner H, Hurley TD. Structure of mitochondrial aldehyde dehydrogenase: the genetic component of ethanol aversion. Structure, 1997; 5: 701-711. [DOI] [PubMed] [Google Scholar]
- 14.Takeshita T, Morimoto K, Mao X, Hashimoto T, Furuyama J. Characterization of the three genotypes of low Km aldehyde dehydrogenase in a Japanese population. Hum Genet, 1994; 94: 217-223. [DOI] [PubMed] [Google Scholar]
- 15.Muramatsu T, Wang ZC, Fang YR, et al. Alcohol and aldehyde dehydrogenase genotypes and drinking behavior of Chinese living in Shanghai. Hum Genet, 1995; 96; 151-154. [DOI] [PubMed] [Google Scholar]
- 16.Yokoyama A, Muramatsu T, Ohmori T, et al. Alcohol-related cancers and aldehyde dehydrogenase-2 in Japanese alcoholics. Carcinogenesis, 1998; 19: 1383-1387. [DOI] [PubMed] [Google Scholar]
- 17.Murata M, Tagawa M, Watanabe S, et al. Genotype difference of aldehyde dehydrogenase 2 gene in alcohol drinkers influences the incidence of Japanese colorectal cancer patients. Jpn J Cancer Res, 1999; 90: 711-719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Takezaki T, Hamajima N, Matsuo K, et al. Association of polymorphisms in the beta-2 and beta-3 adrenoceptor genes with risk of colorectal cancer in Japanese. Int J Clin Oncol, 2001; 6: 117-122. [DOI] [PubMed] [Google Scholar]
- 19.Hamajima N, Matsuo K, Saito T, et al. Interleukin 1 polymorphisms, lifestyle factors, and Helicobacter pylori infection. Jpn J Cancer Res, 2001; 92: 383-389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hamajima N, Saito T, Matsuo K, et al. Polymerase chain reaction with confronting two-pair primers for polymorphism genotyping. Jpn J Cancer Res, 2000; 91: 865-868. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Matsuo K, Hamajima N, Shinoda M, et al. Gene-environment interaction between an aldehyde dehydrogenase-2 (ALDH2) polymorphism and alcohol consumption for the risk of esophageal cancer. Carcinogenesis, 2001; 22: 913-916. [DOI] [PubMed] [Google Scholar]
- 22.Bender R, Lange S. Adjustment for multiple testing - when and how? J Clin Epidemiol, 2001; 54: 343-349. [DOI] [PubMed] [Google Scholar]
- 23.Yokoyama A, Ueno Y, Mizoi Y, Tatsuno Y. Genetic polymorphism of alcohol and aldehyde dehydrogenase and the effects on alcohol metabolism. Jpn J Alcohol Drug Depend, 1993; 28: 13-25. [PubMed] [Google Scholar]
- 24.Vakevainen S, Tillonen J, Agarwal DP, Srivastava N, Salaspuro M. High salivary acetaldehyde after a moderate dose of alcohol in ALDH2-deficient subjects: strong evidence for the local carcinogenic action of acetaldehyde. Alcohol Clin Exp Res, 2000; 24: 873-877. [PubMed] [Google Scholar]
- 25.Kono S, Ikeda M. Correlation between cancer mortality and alcoholic beverage in Japan. Br J Cancer, 1970; 40: 449-455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ogimoto I, Shibata A, Fukuda K. World Cancer Research Fund/American Institute of Cancer Research 1997 recommendations: applicability to digestive tract cancer in Japan. Cancer Cause Control, 2000; 11: 9-23. [DOI] [PubMed] [Google Scholar]
- 27.Hamajima N, Matsuo K, Yuasa H. Adjustment of prognostic effects in prevalent case-control studies on genotype. J Epidemiol, 2001; 11: 204-210, 288. [DOI] [PubMed] [Google Scholar]