Wistar rats were exposed for 2 y r to diesel engine exhaust, carbon black (Printex 90, Degussa, FRC), and ultrafine TiO, (P25, Degussa, FRC) and were subsequently kept in clean air for 6 mo. Particle exposure concentration was increased during the course of the experiment for carbon black and TiO, to reach particle lung loads similar to those found in the diesel soot-exposed rats. The average particle exposure concentrations for diesel soot, carbon black, and JiO, were 7, 11.6, and 70 mg/m', respectively. Lung tumor rates in these rats increased with increasing cumulative particle exposure (mg/m3 x h) independent of the type of particle employed. The exposure to 2.5 mg/m3 diesel soot also induced a significantly increased lung tumor rate, but 0.8 mg/m' diesel soot did not. With this study, it could be demonstrated that the carbon core of diesel soot is mainly responsible for the occurrence of diesel engine exhaust-related lung tumors; the role of diesel soot-attached polycyclic aromatic hydrocarbons (PAH) and NO,-PAH is probably of minor importance in the rat lung. Agglomerates of ultrafine carbon and TiO, particles seem particularly suited to exert toxic effects primarily on alveolar macrophages and alveolar lung particle clearance. Although such lung toxic effects were also seen with the lowest diesel soot exposure concentration (0.8 mg/m3) used, no increased lung tumor rate was detected in this group of rats. Whether this result implies a threshold for the particle-related lung tumor induction mechanism as already discussed by Vostal (7986) or whether the tumor effect was simply not observed because of statistical reasons needs further research on the possible mode of action of ultrafine insoluble particles in the lung. NMRl mice that were kept in the same exposure atmospheres (high diesel soot, car6on black, TiOJ as the rats did not show an increased lung tumor rate. furthermore, there was no treatment-related tumor response in NMRl nor in C57BL/6N mice exposed to diesel exhaust containing 4.5 mg/m3 diesel soot or to the same exhaust dilution but devoid of soot particles. C57BU6N mice were exposed for 24 mo and were subsequently kept in clean air for another 6 mo. Not only the average survival time but also the particle load per gram lung wet weight of the C57BL/6N mice was very similar to rats exposed to 7 mg/m' diesel soot.In 1986, the lung tumor-inducing activity of diesel engine exhaust in rats exposed to diesel soot concentrations of at least 2.2 mg/m3 was observed by various laboratories (Brightwell et al., 1986;Heinrich et al., 1986a;Mauderly et al., 1986). In 1987, based o n these results, diesel exhaust was classified in Germany as a carcinogenic working material. In
