Comparison of biological and chemical predictors of dermal carcinogenicity of petroleum oils

Various methods have been proposed as a basis for cancer hazard labelling of petroleum products. The Institute of Petroleum analytical method termed TP 346/80 was recently recommended to the European Union as the preferred method for this purpose. In this report we compare IP 346/80 with several other predictors of dermal carcinogenicity, including the Mobil PAC Method, the Modified Ames Test (ASTM Method E 1687-95), and the P-32-postlabelling assay for DNA adducts. Oils for assay were selected from a repository of samples previously subjected to mouse skin-painting bioassay. 120 oils were tested in the Modified Ames Assay, 57 by the Mobil PAC Method, 50 by Method IP 346/80, and 48 by the postlabelling procedure. The ability of each assay to distinguish between carcinogenic and noncarcinogenic oils was examined at various suggested discriminators, e.g. Mutagenicity Indices (MI) of 1.0 and 2.0, 3-7 ring PAC contents of 1% and 2% (w/w), and IP 346/80 DMSO-extractables of 2% and 3% (w/w). Various adduct levels were tested for maximum discrimination between carcinogenic and noncarcinogenic oils. The accuracy of each method is reported.