Incremental lifetime cancer risks computed for benzo[a]pyrene and two tobacco-specific N-nitrosamines in mainstream cigarette smoke compared with lung cancer risks derived from epidemiologic data.

The manner in which humans smoke cigarettes is an important determinant of smoking risks. Of the few investigators that have predicted cancer risks from smoking on a chemical-specific basis, most used mainstream cigarette smoke (MCS) carcinogen emissions obtained via machine smoking protocols that only approximate human smoking conditions. Here we use data of Djordjevic et al. [Djordjevic, M.V., Stellman, S.D., Zang, E., 2000. Doses of nicotine and lung carcinogens delivered to cigarette smokers. J. Natl. Cancer Inst. 92, 106-111] for MCS emissions of three carcinogens measured under human smoking conditions to compute probability distributions of incremental lifetime cancer risk (ILCR) values using Monte Carlo simulations. The three carcinogens considered are benzo[a]pyrene, N'-nitrosonornicotine (NNN), and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK). Computed NNK ILCR values were compared with lifetime risks of lung cancer (ILCR(CMD)(obsSigma-lung)) derived from American Cancer Society Cancer Prevention Studies (CPS) I and II. Within the Monte Carlo simulation results, NNK was responsible for the greatest ILCR values for all cancer endpoints: median ILCR values for NNK were approximately 18-fold and 120-fold higher than medians for NNN and benzo[a]pyrene, respectively. For "regular" cigarettes, the NNK median ILCR for lung cancer was lower than ILCR(CMD)(obsSigma-lung) from CPS-I and II by >90-fold for men and >4-fold for women. Given what is known about chemical carcinogens in MCS, this study shows that there is a higher incidence of lung cancer from exposure to MCS than can be predicted with current risk assessment methods using available toxicity and emission data.