CORESTA Meeting, Smoke Science/Product Technology, Stratford-upon-Avon, 2005, SSPTPOST 09

Predictivity and variability of in vitro bioassays routinely used to estimate the toxicity of cigarette smoke

Charles River Laboratories, Montreal, Imperial Tobacco Canada Limited, Montreal, Canada.

An amendment to the Tobacco Act is currently being introduced in Canada. This will require all brands of cigarettes sold in this country to be subjected to a set of three in vitro assays on an annual basis. The purpose of the amendment is to monitor potential changes in the toxicity profile of cigarette smoke over time. The tests will also be used to compare toxicity of standard and reduced ignition propensity cigarettes, which are being introduced at the end of the year as a result of additional legislation. The bioassays are very similar to those used by the In Vitro Task Force in their recent evaluation of relative toxicity of four types of cigarette tobacco, i.e. Ames Bacterial Mutation, Neutral Red Uptake Cytotoxicity and the In Vitro Micronucleus Tests. However, presumably to allow comparability of data between laboratories and across time, the Health Canada methodology is very specific in most aspects of study design. To determine the predictive accuracy of the assays, it is vital to determine the inherent variability within each assay. For example, if the apparent activity of a given brand of cigarettes increases by 25% between testing in 2005 and 2006, is that due to changes in the cigarette, or is it more likely due to the variability of the biological assay. We have examined variability of each bioassay using 2R4F Kentucky reference cigarettes smoked under intense smoking conditions (volume 55 ml; interval: 30 sec.). The influence of the following factors has been examined: Variability between cultures and plates;Replicate smoking; Experiment to experiment variation; Length of time in storage; Activity relative to concurrent positive and negative controls. Coefficients of variance have been calculated for each factor. These results, together with historical control information, should be taken into account when considering the relevance and reliability of the results of individual assays.