Application of toxicological risk assessment principles to the chemical constituents of cigarette smoke.

OBJECTIVE: To provide a hazard prioritisation for reported chemical constituents of cigarette smoke using toxicological risk assessment principles and assumptions. The purpose is to inform prevention efforts using harm reduction. DATA SOURCES: International Agency for Research on Cancer Monographs; California and US Environmental Protection Agency cancer potency factors (CPFs) and reference exposure levels; scientific journals and government reports from the USA, Canada, and New Zealand. STUDY SELECTION: This was an inclusive review of studies reporting yields of cigarette smoke constituents using standard ISO methods. DATA EXTRACTION: Where possible, the midpoint of reported ranges of yields was used. DATA SYNTHESIS: Data on 158 compounds in cigarette smoke were found. Of these, 45 were known or suspected human carcinogens. Cancer potency factors were available for 40 of these compounds and reference exposure levels (RELs) for non-cancer effects were found for 17. A cancer risk index (CRI) was calculated by multiplying yield levels with CPFs. A non-cancer risk index (NCRI) was calculated by dividing yield levels with RELs. Gas phase constituents dominate both CRI and NCRI for cigarette smoke. The contribution of 1,3-butadiene (BDE) to CRI was more than twice that of the next highest contributing carcinogen (acrylonitrile) using potencies from the State of California EPA. Using those potencies from the USEPA, BDE ranked third behind arsenic and acetaldehyde. A comparison of CRI estimates with estimates of smoking related cancer deaths in the USA showed that the CRI underestimates the observed cancer rates by about fivefold using ISO yields in the exposure estimate.
CONCLUSIONS: The application of toxicological risk assessment methods to cigarette smoke provides a plausible and objective framework for the prioritisation of carcinogens and other toxicant hazards in cigarette smoke. However, this framework does not enable the prediction of actual cancer risk for a number of reasons that are discussed. Further, the lack of toxicology data on cardiovascular end points for specific chemicals makes the use of this framework less useful for cardiovascular toxicity. The bases for these priorities need to be constantly re-evaluated as new toxicology information emerges.