Quantitative assessment of a human carcinogenic potency for propylene oxide

The potential for causing carcinogenic and mutagenic effects is the main concern when assessing the risks associated with low-level exposures of humans to the industrially important epoxide, propylene oxide (PO). The available basic information used in estimation of carcinogenic risk has been reviewed. It is concluded that the published data from gavage studies in rodents are less appropriate and that observed cancer incidences from long-term inhalation should preferably be utilized for quantitative risk assessment. Furthermore, PO and ethylene oxide (EO) are directly acting alkylating agents which exhibit several similarities. Although data are less comprehensive for PO than for EO, PO appears to yield a rather uniform alkylation pattern in various tissues. Also, similarly to EO, PO is probably detoxified at a rate which does not vary widely in various mammalian species, including man. For these reasons, the surface-based extrapolation model for estimation of the human equivalent dose may not be appropriate, and the previously derived carcinogenic potency factors should be revised downward. Alternative risk estimates are provided. From the most relevant available studies, we propose a carcinogenic potency factor of 0.001 (mg/kg/day)-1 for PO in humans by inhalation.