Year: 1998
Type: article
Abstract: We thank Bouchardy et al. for bringing to our attention their new data regarding the relationship of alcohol drinking, alcohol dehydrogenase 3 (ADH3) genotype, and oral cancer. However, the limited information they provide makes comparisons between the two studies difficult. Details are missing regarding subjects' age and sex; the anatomic subsites of the tumors; whether incident or prevalent cases were studied; the method used to calculate alcohol intake; the types of alcoholic beverages consumed; the distribution of alcohol intakes by case-control status, by ADH3 genotype, and by disease category for the control subjects; the risks observed with the use of a non-drinker referent group; and the risks associated with each of the following three ADH3 genotypes: ADH1-13 (homozygous for the fast-metabolizing ADH13 allele), ADH1-23 (heterozygous), and ADH2-23 (homozygous for the slow-metabolizing ADH23 allele). In our study (1), we observed an adjusted risk for oral cancer of 1.3 (95% confidence interval [CI] = 0.8-2.4) among subjects with the ADH1-13 genotype compared with subjects with the ADH1-23 or the ADH2-23 genotype, quite similar to the risk observed by Bouchardy et al. (odds ratio [OR] = 1.4; 95% CI = 0.8-2.3). However, Bouchardy et al. did not observe an increased risk associated with the ADH1-13 genotype among consumers with the highest alcohol intakes, in contrast to our study (1). We note that the alcohol-related risk of oral cancer was higher for heavy drinkers (⩾57 drinks/ week) in Puerto Rico (OR = 13.1; 95% CI = 3.9>44.2) (1) than for heavy drinkers (>120 g/day [approximately >70 drinks/week]) in France (OR = 4.8; 95% CI = 2.2–10.7), which may be due to the use of different referent groups (0 drinks/week in our study; ⩽40 g/day [approximately ⩽23 drinks/week] in the study by Bouchardy et al.), differences in alcohol use (i.e., amounts, patterns of use, or beverage types), or misclassification in the exposure assessment. Like Bouchardy et al., we (1) observed no additional risk associated with the ADH3 genotype at intakes associated with risks up to OR = 4.7 (95% CI = 1.6–14.4). A particular concern is that individuals with severe liver disease were excluded from the study by Bouchardy et al. but not from our study (1). The relationships between ADH3 genotype, liver damage, alcohol intake, and ADH activity are unresolved (2–6) and potentially complex. If liver damage and ADH1-13 genotype are positively associated, excluding subjects with severe liver disease will undersample heavy drinkers with the ADH1-13 genotype. While details regarding the exclusion criteria are not given by Bouchardy et al., casecontrol differences could have biased the results toward the null. We interpret the only other study (6) of ADH3 genotype, alcohol intake, and oropharyngeal cancer as suggesting that heavy drinkers with the ADH1-13 genotype may have an elevated risk of upper aerodigestive tract cancers compared with those with the ADH1-23 or the ADH2-23 genotype. Coutelle et al. (6) reported a 2.6-fold (95% CI = 0.710.0) higher risk of oropharyngeal cancer and a 6.1-fold (95% CI = 1.3-28.6) higher risk of laryngeal cancer associated with the ADH1-13 genotype in a small study of French alcoholics (21 oropharyngeal cancer case patients, 18 laryngeal cancer case patients, and 37 control subjects). We agree that further work is needed to clarify the role of genetic susceptibility in the etiology of alcohol-related oral cancer.
Language: English
Subfield: Pathology and Forensic Medicine
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4245744437