The selective determination of potentially carcinogenic polycyclic aromatic compounds in lubricant base oils by the DMSO extraction method IP346 and its correlation to mouse skin painting carcinogenicity assays.

Mineral oils are produced by vacuum distillation of crude oil at temperatures from ∼300 °C to ∼600 °C. Subsequent refining processes to eliminate the carcinogenic potential of mineral oils (by extraction and/or hydrotreatment) are based on the principle of removing substances associated with carcinogenic activity; i.e. PAC (polycyclic aromatic compounds), which include PAH and N or S heterocycles. Traditionally, the carcinogenic potential of the refined product was tested in the mouse skin painting assay. This bioassay is considered the gold standard for petroleum derived products since it uses the most sensitive species and route of exposure, and because mice and humans develop the same type of skin tumors it is a relevant model to assess the carcinogenic potential of mineral oils. Mouse skin painting studies have also been important in distinguishing two types of aromatic compounds found in mineral oil. The first type includes the 3-7 ring PAC associated with potential carcinogenic effects found in the 340-535 °C boiling range, which are removed by refinement. The second type includes highly alkylated aromatic compounds (predominantly 1-2 rings) which are not bioactivated and non-carcinogenic, which are typical of a refined oil. Because mouse skin painting studies are time consuming, a DMSO based method was developed that is capable to distinguish these two types of aromatics. Although this industry method, known as the IP346, has been applied for more than 30 years, the background experimental data underlying its development has not yet been published. This paper presents and discusses the chemical and biological features of mineral oil PAC structures assessed by IP346, especially the crucial role of the DMSO extraction step which allows to discriminate between the two types of aromatics. The DMSO selectivity towards the toxicological relevant PAC is discussed by comparing the composition of the DMSO extract of a distillate aromatic extract and mineral oils of varying viscosities and refining conditions. PAC which have >3 rings (naked or partially alkylated) are preferentially encompassed by the DMSO extract, whereas those PAC which have relatively long alkyl side chains are not. Thus, according to the IP346, refined oils will have lower levels of DMSO extractable material compared to less refined oils. DMSO selectivity towards the potentially carcinogenic >3 ring PAC makes the IP346 method therefore highly correlated to the outcome of mouse skin painting studies, using a pass/fail dichotomy. The accuracy, including the false negative results of the IP346 in the prediction of mineral oil carcinogenicity is discussed. The DMSO based IP346 is thus a simple but clear reflection of refinement efficacy. It links manufacturing conditions to carcinogenic potential of an oil, supported by solid physical-chemical and toxicological associations. In Europe it is the only legally binding method to assess, classify and label lubricating base oils and inherently more reliable for hazard assessment than the determination of an arbitrary selection of PAH.