The 13th World Conference on Tobacco OR Health

Assessing hazards of tobacco product constituents and emissions pose unique challenges because of their very complex chemical nature and the limited availability of validated models to predict human toxicity. Various in vitro models can address specific endpoints, can be used to rank responses of individual chemicals within defined classes, and can yield useful mechanistic information. However, induction of adverse effects leading to endpoints such as cancer, genetic disease, reproductive disorders and immunotoxicity, involves complex biological interactions including multicellular, multiorgan, hormonal, neural, vascular and immunological systems. Modelling of such complex adverse effects cannot be accomplished by the use of non-animal tests. Experimental inhalation carcinogenicity models are poor predictors for human lung cancer. The Salmonella mutagenicity test is the most predictive of the genetic toxicity tests for animal carcinogenicity, but there is very weak relationship between mutagenic potency and quantitative rodent carcinogenicity. However, for single chemicals there is a very good correlation between carcinogenic and in vivo genotoxic potency estimates. A toxicological risk index assessment framework ranks chemical hazards in cigarette smoke, both for cancer and non-cancer effects. This framework may be used for prioritising such compounds in terms of their contribution to toxicity of cigarette smoke. There are limitations of extrapolating results from experimental models to humans, and difficulties in animal models to simulate the conditions when humans are exposed to tobacco product constituents and emissions. Thus, the way forward for assessing hazards and risks related to tobacco product use will presumably be to develop and use predictive human effect biomarkers.