The Health Outcome
Alcohol dehydrogenase ( ADH ) and aldehyde dehydrogenase (ALDH ) are key enzymes involved in ethanol metabolism. Ethanol is first metabolized by ADH to acetaldehyde, which is then converted into acetate by ALDH. The ALDH2*2 and ADH2*2 alleles are prevalent genetic polymorphisms in East Asians. Among Japanese, Chinese, and Koreans, the prevalence of ALDH2*2 (heterozygous and homozygous) is 30–50% while the polymorphism is barely found in other races and ethnicities (1). Both heterozygous and homozygous genotypes for ALDH2*2 lead to functionally inactive forms of ALDH2. The ADH2*2 allele encodes super-active forms of ADH2, compared with the enzyme encoded by ADH2*1, a slow metabolizer.
The ALDH2*2 allele causes severe acetaldehydemia, which produces the alcohol flushing response, characterized by facial flushing, tachycardia and drowsiness. These unpleasant symptoms tend to deter people from drinking. Thus, it has been said that inactive ALDH2 is a “protective ” factor against developing alcoholism (2).
At the same time, ALDH2*2 has been shown to be a risk factor for GI tract cancers. Acetaldehyde has been shown to be carcinogenic in in vitro and in vivo animal studies (2). Thus, it is reasonable to hypothesize that the build-up of acetaldehyde caused by inactive ALDH2*2 may contribute to carcinogenesis in epithelial cells of the esophagus. Smoking and alcohol are strong risk factors for squamous cell esophageal cancer, and ALDH2*2 has been consistently associated with a higher risk of esophageal cancer among alcoholics.
ADH2*1, the slow-metabolizing polymorphic form of ADH2, has also been associated with increased risk of esophageal cancer; this finding contrasts with the expected effect based on function, because the fast form of ADH2 acts to maintain higher levels of acetaldehyde, the presumed carcinogen. Among alcoholics, the multiplicative in teraction between ALDH*2 and ADH2*1 was found to increase risk of esophageal cancer (2).
Cancer is a leading cause of mortality in Japan, where more than 10,000 patients are diagnosed each year with esophageal cancer(3) . The 5 year survival rate is about 40%, which is considerably higher than in the U.S, but when detected early, the 5 year survival markedly improves to 80 to 100% (3,4). Thus, identifying a high-risk group that might benefit most from endoscopic screening is an important public health strategy for preventing this cancer.
One currently used approach is to detect those with “flushing response” by using an “ethanol patch test,” which detects cutaneous flushing after dermal exposure to alcohol. The problem with this test is that it only detects a current flushing response. It has been shown that people with inactive ALDH2 can develop tolerance to the alcohol flushing response after habitual drinking (5). Thus, many heavy drinkers with inactive ALDH2 will not present with an active flushing response, although they are a group at increased risk of developing esophageal cancer.
The Finding
In a previous publication, Yokoyama et al conducted a hospital based case-control study examining various metabolic enzyme genotypes in relation to esophageal cancer risk in Japanese men. The cases were 234 male Japanese esophageal cancer patients from 4 urban hospitals; 634 Japanese men registered for annual physical exam at 2 urban clinics were enrolled as controls (2). The investigators performed RFLP genotyping of metabolic enzymes and collected self-reported information on diet, smoking, and drinking habits.
The authors found a multiplicative in teraction between ALDH2*2 and ADH2*1 related to risk of esophageal cancer and noted that this finding was consistent with studies done in alcoholics (2). Inactive ALDH2 was associated with increased risk of esophageal cancer in light to moderate drinkers as well as in heavy drinkers. Population attributable risks were also reported: 90% for alcohol overall (regardless of genotype ), and 68.5% for alcohol in the inactive ALDH2 group.
In the current study, the authors used data from participants in the previous case-control study to examine the reliability of a simple questionnaire for detecting inactive ALDH2 and its clinical validity in predicting risk for esophageal cancer. A simple questionnaire was administered to ask about current and past flushing responses after drinking. To assess the analytical validity of the questionnaire, results were compared with genotyping data from the previous study, which was used as the gold standard.
They found that specificity and sensitivity of the questionnaire for the cases were 82.3% and 84.8% respectively and for the controls were 88% and 90.1% respectively(5). When sub-analysis was done to evaluate the difference in sensitivity and specificity between the cases and the controls, the authors found that ADH2 polymorphism modified the flushing response. Those with ADH2*1 had a lower sensitivity and specificity compared to those with ADH2*2. After adjusting for the ADH2 polymorphism, the sensitivity and specificity became comparable between the cases and the controls. Thus the validity of the questionnaire was weaker for those with ADH1*1 genotype. Since there is no alternative to a genetic analysis for detecting ADH2*1, the utility of this questionnaire for assessing risk of esophageal cancer is limited because it the results can only be used to infer inactive ALDH2; no alternative to genotyping currently exists for detecting ADH2*1.
A multiple logistic regression was fitted to predict risk of esophageal cancer. Combined categories of the flushing response (in lieu of the metabolic enzyme genotypes) and alcohol drinking categories were analyzed, using the never/rare drinking category as the referent group. The ORs were adjusted for age, frequency of “drinking strong alcohol beverage straight ”( not well described ), smoking ( in pack-years ), and intake frequency of vegetables and fruits.
The authors found that flushing was associated with higher risk of esophageal cancer in light, moderate, and heavy drinking categories (OR = 5.3, 4.2, and 4.7 ), even though the strength of association was slightly attenuated compared with results using genotypes in the model (5).
The authors concluded that this simple questionnaire can reliably detect inactive ALDH2 because it asks about former as well as current flushing status, and that it can reliably predict increased risk for esophageal cancer. They suggested that the simple questionnaire may be a useful surrogate marker for inactive ALDH2 when conducting large-scale epidemiologic studies, avoiding the need for costly genetic analysis. In addition, the flushing status questionnaire could be a cost-effective tool in public health education for drinkers and clinicians alike in identifying high-risk individuals for esophageal cancer.
Public Health Implications
The study was done among Japanese men over 40 years old. Studying this homogeneous population avoids the potential for population stratification and seems like a relevant choice, not only because the genotype is found mostly in East Asians, but because of the cultural/ social importance of drinking in Japanese men.
Choosing controls from men who came for an annual physical exam seems reasonable because those diagnosed with esophageal cancer in this population would probably be referred to the same urban cancer centers for treatment. Selecting controls from a single city (while cases were selected from hospitals in 3 different cities) seems to be acceptable, considering the homogeneity of this population in terms of genotypes, dietary, smoking, and drinking habits, and socioeconomic factors. In fact, the authors cite the results from a National Nutrition Survey to make this point.
However, differences in participation rate (almost 99% in cases compared with 83% in controls) and age distribution between cases and controls may suggest potential selection bias. In addition, analyzing self-reported flushing status allows greater room for recall bias: considering that habitual drinking leads to tolerance, an unpleasant response for a year or two could be forgotten after many years with no problem. In fact, some people responded “unknown,” and were classified as non-flushing. Finally, the small numbers in certain exposure categories may make the estimates unstable (as reflected in the width of confidence intervals).
It is also worrisome that the sensitivity and specificity of the questionnaire for ALDH2 genotype differed between the cases and the controls, and that the questionnaire appeared to predict ALDH2 genotype but not the other well-known genetic modifier of the flushing response, ADH2. It would be interesting to see the reliability of this questionnaire in detecting the ALDH2 genotype in a different study population, perhaps a larger cohort of healthy patients that even include women. Since most of the studies that found an association between ALDH2 and esophageal cancer were done in alcoholics, more studies are needed to verify the findings in this study for light and moderate drinkers.
Both the ALDH2*2 allele and alcohol use are quite prevalent among Japanese men, making this association, if true, of considerable interest. A simple questionnaire that can reliably detect a high risk genotype could be an inexpensive screening tool for identifying a high-risk population that could benefit from early screening and public health education to change drinking behavior.
References
- Brennan P, et al. Pooled Analysis of Alcohol Dehydrogenase Genotypes and Head and Neck Cancer: HuGe Review. Am J Epidemiol. 2004 Jan;159 (1):1-16.
- Yokoyama T, et al. Genetic polymorphisms of alcohol and aldehyde dehydrogenases and glutathione S-transferase M1 and drinking, smoking, and diet in Japanese men with esophageal squamous cell carcinoma. Carcinogenesis. 2002; 23 ( 11 ): 1851-1859.
- Japan National Cancer Center Homepage (Japanese pages)
- Tierney L, et al, ed. Current Medical Diagnosis and Treatment 2003. 42 nd ed. Lange Medical books/McGraw-Hill. 2003.
- Yokoyama T, et al. Alcohol Flashing, Alcohol and Aldehyde Dehydrogenase Genotypes, and Risk for Esophageal Squamous Cell Carcinoma in Japanese Men. CEBP. 2003 Nov; 12: 1227-1233.
