Risk assessment in the regulatory science process.

Lars E. Rutqvist, M.D., Ph.D,

Other Contributors:

October 28, 2010

The approach by Ayo-Yusuf and Connolly (2010) to evaluate cancer risks of smokeless tobacco products (STP) addresses issues that could be relevant to modified risk claims for Swedish snus tobacco products. We disagree with the authors' conclusions, and in some cases they simply have the facts wrong. Nonetheless, the issues presented warrant consideration by the tobacco science community, including the FDA Center for Tobacco Products and the newly formed Institute of Medicine committee on modified risk.

The authors' use of the EPA risk assessment process is intriguing. In the hands of someone familiar with the process it may have resulted in a useful contribution to the scientific literature. Unfortunately the authors do not seem to understand the process because they did not follow the well documented EPA guidance.

Recently a National Academy of Sciences committee issued a report examining the EPA risk assessment process (Science and Decisions: Advancing Risk Assessment, 2009). The committee cited the importance of making risk assessments data driven, and when data are unavailable ensuring that default assumptions are accurately characterized. When there is an abundance of information - particularly epidemiological data - the credibility and utility of the risk assessment increases significantly.

There is extensive health-related information available for Swedish snus. That is why an EPA risk assessment approach is worth pursuing. However, the authors did not use all the available data, particularly not the epidemiological documentation. EPA scientists have some flexibility in choosing "critical" studies, but they must justify their decisions, and use "all the relevant and available scientific information." EPA scientists cannot disregard data simply because of personal biases.

Also, instead of comparing life-time use of STP with only 12-weeks medicinal nicotine, the authors should have made a comparison with long- term cigarette smoking. This could then be expanded to comparisons with selected smoke components. It would be enlightening, for example, to determine risk assessment values for the carcinogenic potential of such components as 1, 3-butadiene compared with the authors' claims for tobacco -specific nitrosamines and benzo(a)pyrene.

In summary, we support the application of risk assessment in the regulatory science process. However, if the EPA risk assessment model is used to evaluate potential modified risk tobacco products, it should be done so by scientists experienced in the EPA process who have an understanding and appreciation of the available data.

References: Ayo-Yusuf, O., A., Connolly, G., N. Applying toxicological risk assessment principles to constituents of smokeless tobacco products: implications for product regulation. Tob Control, published online Oct 5, 2010. doi: 10.1136/tc.2010.037135

Science and Decisions: Advancing Risk Assessment. Committee on Improving Risk Analysis Approaches Used by the U.S. EPA. Board on Environmental Studies and Toxicology. Division on Earth and Life Studies. National Research Council of the National Academies. National Academic Press, Washington D.C., 2009.

Conflict of Interest:

Lars E. Rutqvist is employed by Swedish Match AB. Chris Coggins acts in a consulting role to the company.

Conflict of Interest

None declared