CYP2A6 polymorphism, nicotine, and environmental nitrosamines

Abstract

We reported (S J London and colleagues, March 13, p 898)1London SJ Idle JR Daly AK Coetzee GA Genetic variation of CYP2A6, smoking and risk of cancer.Lancet. 1999; 353: 898-899Summary Full Text Full Text PDF PubMed Scopus (128) Google Scholar that polymorphism in the CYP2A6 gene, which codes for a cytochrome P450 isozyme that metabolises both nicotine and nitrosamines, seemed to have little influence on the propensity either to smoke cigarettes or to develop lung cancer. I have two differences of opinion with my co-authors that were not debated before submission of this research letter. First, ours was a genetic epidemiological investigation of risk factors for lung cancer and was not designed to detect factors that might affect acceptability of, or aversion, to nicotine in tobacco products. However, we did undertake CYP2A6 genotyping of lung cancer cases and controls, with known smoking histories. Unfortunately, our definition of an ever smoker was someone with a lifetime smoking history of 100 cigarettes or more. Our first hypothesis, and that of a group from Toronto,2Pianezza ML Sellers EM Tyndale RF Nicotine metabolism defect reduces smoking.Nature. 1998; 393: 750Crossref PubMed Scopus (375) Google Scholar was that people with one or two inactivating CYP2A6 alleles, who would have a lowered capacity to detoxicate metabolically the nicotine absorbed during cigarette smoking, might have an aversion to nicotine, and thus choose not to smoke. I judge that our classification of ever smokers and never smokers has insufficient capability to detect people who might have a genetic and metabolically based intolerability to cigarettes. Second, we designed the study to examine the potential relation between CYP2A6 genotypes and lung cancer causation. Two (1%) of 182 cases and 11 (2%) of 460 controls had the metabolically compromised genotype (two inactivating alleles), generating a non-significant odds ratio for lung cancer of 0·5 in the combined smokers and non-smokers and 0·4 for smokers alone, a finding which, if statistically significant, would have been commensurate with the second hypothesis that CYP2A6-deficient individuals have an impaired metabolic activation of nitrosamines. The question is which nitrosamines? There is little reason to suppose that these must be the so-called tobacco-specific nitrosamines (TSN). Orally administered diethylnitrosamine (not a TSN), for example, which caused tracheobronchial tumours in hamsters3Herrald KMcD Dunham LJ Induction of tumors in the Syrian hamster with diethylnitrosamine (N-nitrosodiethylamine).Cancer Res. 1963; 23: 773-777PubMed Google Scholar is abundant in bacon, cheese, and fish, and microgram amounts of such dietary nitrosamines are much greater than those obtainable from cigarette smoking. I have yet to see the experimental evidence that smoking is the major cause of lung cancer and I am not aware that the burden of epidemiological evidence can refute an important role of diet. We can now formulate a new pharmacogenetic hypothesis for lung cancer: that persons who inherit the greatest CYP2A6 metabolic capacity will be those with the fastest metabolism of nicotine, and therefore those most likely to smoke cigarettes, those most likely to activate environmental nitrosamines, particularly from the diet, and, those most likely to develop lung cancer, not contingent on them being an ever smoker. As with traditional epidemiology, the pitfalls of confounders will need to be avoided (other genotypes and environmental and lifestyle factors). Finally, when a metabolic phenotype is a causative factor, misclassification of individuals by incomplete genotyping will drive the odds ratio towards unity. As yet undiscovered CYP2A6 inactivating alleles probably exist in certain human populations.4Nunoya K Yokoi T Kimura K et al.A new deleted allele in the human cytochrome P450 2A6 (CYP2A6) gene found in individuals showing poor metabolic capacity to coumarin and (+)-cis-3,5-dimethyl-2-(3-pyridyl)thiazolidin-4-one hydrochloride (SM-12502).Pharmacogenetics. 1998; 8: 239-249Crossref PubMed Scopus (112) Google Scholar