Letter to the Editor

To the Editor: With interest I read the paper by Kuempel and colleagues (2006) on lung dosimetry and risk assessment of nanoparticles. I'd like to congratulate the authors for this clear exposition of the quantitative risk assessment (QRA) steps for estimating human-equivalent critical exposures using rat dose-response data. However, the accuracy (precision and validity) of the approach is still to be assessed and the authors missed presenting a strategy on how this could be achieved. Thus, I'd like to make a suggestion. One approach to increase our belief in QRA is to show that estimates can be transferred reliably between different animal species. It would be helpful if the authors could show how accurately rat-based critical dose estimates or lung doses can be transferred by QRA to mice or hamsters (and vice versa). Some study results on mice and hamsters are available for validation (Heinrich et al., 1995; Bermudez et al., 2004; Nikula et al., 1995). Such comparison tests across different animal species are urgently needed because, first, QRA combines independent parts of knowledge of limited precision (which decreases the overall precision of the approach), and second, we are aware of examples where the extrapolation from single-species data to humans can be grossly misleading (Kelsey, 1963; Zurlo et al., 2002) (which may shed doubt on the validity of a QRA generating human-equivalent exposure estimates based on rat data only).

I would like to add that the critical dose estimate (QRA step 2) may depend remarkably on the decision whether a threshold parameter is included in the model or not (against the unproven paradigm of linearity in the low risk region, cf. footnote on p. 718) and that it may be important for a reliable analysis to adjust for differential survival between the rats (Morfeld et al., 2006).