AIM: To evaluate the association between genetic polymorphisms in CYP2E1, ALDH2 and ADH1B and the risk of esophageal squamous cell carcinoma (ESCC) in a high risk area of Gansu Province, in Chinese males.

METHODS: A case-control study was conducted to investigate the genetic polymorphisms of these enzymes (CYP2E1*c1/*c2, ALDH2*1/*2 and ADH1B *1/*1 genotypes). A total of 80 esophageal cancer cases and 480 controls were recruited.

RESULTS: Compared with controls, cases had a greater prevalence of heavier alcohol consumption (53.8% vs 16.2%) and a higher proportion of alcohol drinkers with > 30 drink-years (28.8% vs 13.5%). Heavier alcohol consumption and alcohol drinking with > 30 drink-years increased the risk of ESCC, with ORs (95% CI) of 3.20 (1.32-9.65) and 1.68 (0.96-3.21). CYP2E1 (*c1/*c1), ALDH2 (*1/*2) and ADH1B (*1/*1) genotype frequencies were higher among patients with squamous cell carcinomas, at a level close to statistical significance (P = 0.014; P = 0.094; P = 0.0001 respectively). There were synergistic interactions among alcohol drinking and ALDH2, ADH1B and CYP2E1 genotypes. The risk of the ESCC in moderate-to-heavy drinkers with an inactive ALDH2 encoded by ALDH2*1/*2 as well as ADH1B encoded by ADH1B *1/*1 and CYP2E1 encoded by CYP2E1 *c1/*c1 was higher than that in the never/rare-to-light drinkers with an active ALDH2 (*1/*1 genotype) as well as ADH1B (*1/*2 + *2/*2) and CYP2E1 (*c1/*c2 + *c2/*c2) genotypes, with a statistically significant difference; ORs (95% CI) of 8.58 (3.28-22.68), 27.12 (8.52-70.19) and 7.64 (2.82-11.31) respectively. The risk of the ESCC in moderate-to-heavy drinkers with ALDH2 (*1/*2) combined the ADH1B (*1/*1) genotype or ALDH2 (*1/*2) combined the CYP2E1 (*c1/*c1) genotype leads to synergistic interactions, higher than drinkers with ALDH2 (*1/*1) + ADH1B (*1/*2 + *2/*2), ALDH2 (*1/*1) + CYP2E1 (*c1/*c2 + *c2/*c2) respectively , ORs (95% CI) of 7.46 (3.28-18.32) and 6.82 (1.44-9.76) respectively. Individuals with the ADH1B combined the CYP2E1 genotype showed no synergistic interaction.

CONCLUSION: In our study, we found that alcohol consumption and polymorphisms in the CYP2E1, ADH1B and ALDH2 genes are important risk factors for ESCC, and that there was a synergistic interaction among polymorphisms in the CYP2E1, ALDH2 and ADH1B genes and heavy alcohol drinking, in Chinese males living in Gansu Province, China.

Esophageal carcinoma is the seventh leading cause of cancer deaths worldwide[1]. It is a devastating disease with very few patients cured once diagnosed. Esophageal squamous cell carcinoma (ESCC) is one of the most common malignancies in Gansu province, China. There is great geographic variation in the occurrence of this tumor type in China, including exceptionally high-risk areas such as Gansu Province in the Northwest of China. Within high-risk regions in China, there is a strong tendency toward alcohol drinking, suggesting that genetic susceptibility, in conjunction with alcohol consumption, plays a role in the etiology of ESCC.

Epidemiologic studies have demonstrated that drinking alcoholic beverages is causally related to the development of ESCC[23]. Genetic polymorphisms in the genes encoding cytochrome P4502E1 (CYP2E1)[4–7], aldehyde dehydrogenase-2 (ALDH2) and alcohol dehydrogenase-1B (ADH1B; previously called ADH2) affect the metabolism of alcohol[8–12]. There have been some studies on the roles of alcohol and polymorphisms in the CYP2E1, ALDH2 and ADH2 genes in ESCC[13–15]. However, their results were conflicting.

To provide further data on this issue, we evaluated the susceptibility to ESCC conferred by CYP2E1, ALDH2 and ADH1B genetic polymorphisms, and defined the individual and combined roles of these genes and alcohol consumption in a high risk area for ESCC in Chinese males.

The case participants in this study were 80 Gansu males with ESCC treated at the Department of Gastroenterology, First Hospital of Lanzhou University and the Department of Gastroenterology, Affiliated Hospital of Gansu College of Traditional Chinese Medicine. All were registered between September 2004 and March 2007.

The 480 age-and-gender matching controls consisted of cancer-free men who received annual health checkups at two Lanzhou clinics between September 2004 and March 2007. Gansu was the ancestral home for all.

Each participant independently completed a structured questionnaire concerning his alcohol drinking habits, and those with cancer were instructed to report their habits before they were diagnosed with cancer. Each participant was asked to classify himself as a drinker or non-drinker, and to report alcohol intake as the frequency of consumption and usual amount(s) and type(s) of alcoholic beverage(s) consumed. The subjects were classified as never drinks alcohol, or drinkers who consumed 200 g/wk (light drinkers), 200-400 g/wk (moderate drinkers) or 400+ g/wk (heavy drinkers).

Polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) methods or a PCR method were performed on lymphocyte DNA samples from all participants, without any knowledge of their cancer status, to determine their P4502E1, ALDH2 and ADH1B genotypes (Tables 1 and 2).

The Rsa I and Pst I linkage disequilibrium poly-morphisms of the P4502E1 gene were determined according to the method of Hayashi et al[16] with some modification. The final mixture (50 &mgr;L) was prepared containing 0.3 &mgr;g of DNA, 0.2 &mgr;mol/L of dNTP, 0.3 &mgr;mol/L of each of the primers, 1.5 &mgr;mol/L of MgCl₂, 5 &mgr;L of 10 × buffer, and 2 U of Taq polymerase. Briefly, the samples were denatured at 94°C for 2 min and submitted to 40 cycles of 1 min at 94°C (denaturation), 50 s at 50°C (annealing) and 50 s at 72°C (extension), with a final extension at 72°C for 10 min. PCR products (15 &mgr;L) were digested by Pst I or Rsa I restriction enzymes (1 &mgr;L of a 10 U/&mgr;L preparation) for 18 h at 37°C. Fragments were separated on 40 g/L low melting point agarose gels, and stained with ethidium bromide.

ALDH2 and ADH1B[17] polymorphisms were determined by PCR-RFLP as previously described. Each PCR analysis was performed twice, double blindly.

The allele frequency was determined by direct counting. Deviation of the genotype distribution from Hardy-Weinberg equilibrium was analyzed by the exact test. Fisher’s exact test was used for comparing group statistics. The Spearman rank-correlation analysis was used as a nonparametric test for trend. All P-values were obtained from 2-sided tests. Associations between genotypes or other potential risk factors and ESCC are expressed as odds ratios (ORs) with 95% confidence intervals (CIs) adjusted for the effects of several possible confounders using a multiple logistic regression model and the STEPWISE method.

Five hundred and sixty males were enrolled in this study. The cancer cases were age-and-gender matched with cancer-free control subjects. The mean age of the patients was 60.2 ± 8.9, ranging from 49 to 75 years of age. The mean age of the controls was 59.7 ± 9.7, ranging from 49 to 73 years of age.

Table 3 shows alcohol drinking in esophageal cancer cases and control subjects. We observed that, compared with controls, cases had a greater prevalence of heavier alcohol consumption (53.8% vs 16.2%) and a higher proportion of alcohol drinkers with > 30 drink-years (28.8% vs 13.5%). Heavier alcohol consumption and alcohol drinking with > 30 drink-years increased the risk of ESCC, with ORs (95% CI) of 3.20 (1.32-9.65) and 1.68 (0.96-3.21).

Table 4 shows the distributions of ALDH2, ADH1B and CYP2E1 genotypes. Cases and controls differed significantly in the distributions of these genotypes. These genotypes significantly deviated from the Hardy-Weinberg equilibrium (HWE) in ESCC cases, but among controls, all genotypes were in HWE.

There are two ALDH2 alleles (*1 and *2) and three genotypes: *1/*1 (GG, typical homozygote), *1/*2 (GA, heterozygote) and *2/*2 (AA, atypical homozygote), and the distributions of these genotypes were significantly different between the esophageal cancer group and the control group (χ² = 2.89, P < 0.1). The prevalence of the inactive ALDH2 encoded by ALDH2*1/*2 was correlated with susceptibility to esophageal cancer, with an OR (95% CI) of 2.89 (1.11-5.64).

There are three ADH1B genotypes: The wild-type genotype (*1/*1), heterozygote (*1/*2) and homozygote (*2/*2) genotypes. The prevalence of the less-active ADH1B encoded by ADH1B *1/*1 increase the risk of esophageal cancer. (χ² = 18.664, P < 0.0001), OR (95% CI) of 3.67 (1.26-8.73).

There are three CYP2E1 genotypes: wild homozygote (*c1/*c1), heterozygote (*c1/*c2) and mutated homozygote (*c2/*c2) genotypes. A significant difference in the distributions of the three Pst I/Rsa I genotypes of CYP2E1 was found between the esophageal cancer group and the control group (χ² = 5.977, P < 0.05). The c1/c1 genotype was correlated with susceptibility to esophageal cancer, with an OR (95% CI) of 2.82 (1.23-6.55).

Tables 5 and 6 show the frequency distributions and ORs for each combination of alcohol drinking habits and ALDH2, ADH1B and CYP2E1 genotypes.

The risk of ESCC was 8.58-fold higher in moderate-to-heavy drinkers with inactive ALDH2 (encoded by ALDH2*1/*2) than in the never/rare-to-light drinkers with an active ALDH2 (encoded by ALDH2*1/*1).

When the ORs were compared within each alcohol drinking category, the risk of the ESCC associated with ALDH2 (*1/*2) versus ALDH2 (*1/*1) was 3.12-fold greater among moderate-to-heavy drinkers, whereas no significant increase in risk was observed with ALDH2 *1/*2 versus ALDH2 *1/*1 among never/rare-to-light drinkers.

The results for the ADH1B genotype showed that only among moderate-to-heavy drinkers did the less-active ADH1B *1/*1 genotype significantly increase the risk of cancer, with an OR (95% CI) 5.48 (1.98-14.55). Thus, the risk of the cancer in moderate-to-heavy drinkers with the ADH1B (1/1) genotype was markedly higher (by 27.12-fold) than that in never/rare-to-light drinkers with a super-active ADH1B encoded by ADH1B *1/*2 and ADH1B *2/*2.

The results for the CYP2E1 genotype showed that, among moderate-to-heavy drinkers, the CYP2E1 (*c1/*c1) genotype significantly increased the risk of cancer, with an OR (95% CI) of 5.32 (1.98-14.55). Thus, the risk of the cancer in moderate-to-heavy drinkers with CYP2E1 (*c1/*c1) was markedly higher (by 7.64-fold) than that in never/rare-to-light drinkers with a super-active CYP2E1 (genotypes *c1/*c2 + *c2/*c2).

The results of our analysis of ALDH2 and ADH1B genotypes showed that among moderate-to-heavy drinkers, the combination of ALDH2 (*1/*2) and ADH1B (*1/*1) genes synergistically increased the risk of cancer, to 7.46-fold higher than that in drinkers with an ALDH2 (*1/*1) + ADH1B (*1/*2 + *2/*2) genotype.

The results of an analysis of the effects of combinations of ALDH2 and CYP2E1 genotypes showed that among moderate-to-heavy drinkers, the combination of ALDH2 (*1/*2) and CYP2E1 (*c1/*c1) genes synergistically increased the risk of cancer, to 6.82-fold higher than that in drinkers with an ALDH2 (*1/*1) + CYP2E1 (*c1/*c2 + *c2/*c2) genotype.

There was no combinatorial effect of ADH1B and CYP2E1 genotypes.

In the present study, we examined the associations of ESCC with ALDH2, ADH2 and CYP2E1 genetic polymorphisms in conjunction with alcohol drinking habits among a population at high risk of esophageal cancer. The study was conducted in Gansu province, an area of China with a relatively high alcohol consumption rate.

We found that alcohol intake was associated with ESCC, and that polymorphisms in CYP2E1, as well as in the genes encoding alcohol and aldehyde dehydrogenases (ADH1B and ALDH2) are important risk factors for ESCC in Chinese men living in this high risk area.

We found that heavier alcohol consumption and alcohol drinking for > 30 drink-years could increase the risk of ESCC, with ORs (95% CI) of 3.20 (1.32-9.65) and 1.68 (0.96-3.21), respectively. There is substantial evidence that drinking alcohol increases the risk of ESCC. Alcohol can be considered to induce DNA damage and result in the modification of nucleotides. Our risk estimates were consistent with those of previous studies[18]. This study confirms that alcohol consumption contributes to the etiology of ESCC.

It has been reported that individual differences in cytochrome P450 (CYP) gene expression may contribute to a person’s individual susceptibility to pro-carcinogens and, subsequently, to the development of malignancies. CYP plays a central role in the metabolism of carcinogens by activating oxidation reactions, and may be expressed in esophageal mucosa.

CYP2E1 metabolizes ethanol to acetaldehyde and is primarily responsible for the metabolic activation of many low molecular weight carcinogens, including certain nitrosamines, which may be involved in carcinogenesis of the esophagus. This enzyme is also believed to participate in the oxidation of other compounds, such as ethanol, to produce reactive free radicals that may initiate lipid peroxidation and consequently influence carcinogenesis. In addition, it effectively reduces dioxygen to give rise to radical species, thus contributing to lipid peroxidation and oxidative inhibition. Individuals with the variant Pst I/Rsa I allele (*c1/*c2 or *c2/*c2) have a lower basal CYP2E1 activity. Two studies found an association between the CYP2E1 Pst I /Rsa I variant allele and a decreased risk of ESCC[19]. The results of our study indicated that the CYP2E1 *c1/*c1 or c1 allele increased the susceptibility to ESCC in a Gansu population (P = 0.014), and that there are synergistic interactions between CYP2E1 (*c1/*c1) and alcohol drinking; the risk of ESCC in moderate-to-heavy drinkers with the CYP2E1 (*c1/*c1) genotype was markedly higher, by 7.64-fold, than in never/rare-to-light drinkers with a super-active CYP2E1 (genotypes *c1/*c2 + *c2/*c2).

The genetic polymorphisms of alcohol and aldehyde dehydrogenases affect the metabolism of alcohol. Aldehyde dehydrogenase-2 (ALDH2) is the key enzyme for elimination of acetaldehyde. Acetaldehyde, a recognized animal carcinogen derived from alcohol, may play an important role in the pathogenesis of alcohol-related cancers, such as esophageal cancer. In persons with inactive ALDH2 encoded by ALDH2*1/*2, the body fails to metabolize acetaldehyde rapidly, leading to excessive accumulation of acetaldehyde.

Polymorphisms in the ADH1B and ALDH2 genes associated with the risk of esophageal cancer have been described in several studies[20–24]. ADH1B (*1/*1) has also been demonstrated to enhance the risk of esophageal cancer, and the combination of ALDH2 (*1/*2) and ADH1B (*1/*1) increased the risk of esophageal cancer[8].

In our study, we found that polymorphisms in the genes encoding alcohol and aldehyde dehydrogenases (ADH1B and ALDH2) are important risk factors for ESCC in a Gansu population, and individuals with the ADH1B (*1/*1) genotype had a 5.32-fold risk (1.98-14.55) of developing esophageal cancer compared with those with the ADH1B (*2/*2) genotype.

Individuals with the ALDH2 (*1/*2) genotype had a 3.12-fold higher risk (1.86-6.58) of developing esophageal cancer than those with the ALDH2 (*1/*1) genotype, among male Chinese moderate-to-heavy drinkers. We also found that individuals with a combined ALDH2 (*1/*2) and ADH1B (*1/*1) genotype showed a dramatically increased risk of ESCC, with an OR (95% CI) 7.46 (3.28-18.32), which is higher than those due to the respective genotypes. These findings indicate the ALDH2 (*1/*2) genotype has synergistic interactions with the ADH1B (*1/*1) genotype, contributing to the development of ESCC. Our study confirmed the findings of Tetsuji[8].

The significant finding in this study was the interaction between the CYP2E1 and ALDH2 genotypes and heavy alcohol drinking, using a case-control study design. Previous studies have not examined this issue in detail and, to our knowledge, this is the first study to show a significant interaction between the CYP2E1 and ALDH2 genotypes and alcohol drinking. In our study, we found synergistic interactions among polymorphisms in the CYP2E1 and ALDH2 genotypes and heavy alcohol drinking; individuals with a combined ALDH2 (*1/*2) and CYP2E1 (*c1/*c1) genotype showed a dramatically increased risk of ESCC, with an OR (95% CI) of 6.82 (1.44-9.76), which is higher than those due to the respective genotypes.

Conflicting results from studies[25–28] in different countries and areas show the complexity of the biological mechanisms underlying ESCC. The susceptibility to ESCC may be correlated with genes, environment, area, race or other factors. The results of this study demonstrate that CYP2E1 (*c1/*c1), ALDH2 (*1/*2) and ADH1B (*1/*1) genotypes are associated with esophageal cancer risk among moderate-to-heavy drinking Chinese males in Gansu province. In addition, this study also showed ALDH2 (*1/*2) and CYP2E1 (*c1/*c1) carriers, and ALDH2 (*1/*2) and ADH1B (*1/*1) carriers have a much higher risk of developing esophageal cancer, especially among alcohol drinkers. Future studies are needed to examine the biological mechanisms involved, and to evaluate the contribution of gene and environment interactions to the risk of ESCC.