In contrast to BBDR models, the NRC vision included a central role for TP models that quantitatively model the cellular responses in terms of more basic biochemical and genomic responses.
Most of the problems described above that complicate efforts to use BBDR to estimate low-dose risk will be present in TP models.
In fact, having such data may be necessary to fully implement a TP model, just as data on the apical response have been used in the development of BBDR models for in vivo responses, (e.
CONCLUSIONS: The problems discussed here appear so intractable that we conclude that BBDR models are unlikely to be fruitful in reducing uncertainty in quantitative estimates of human risk from low-level exposures in the foreseeable future.
BBDR models are predictive models that describe biological processes at the cellular and molecular level to link external exposure to an adverse apical response.
Difficulties inherent in BBDR models limit the models' ability to provide reliable estimates of low-dose risk in humans.
BBDR modeling has the same problems for estimating low-dose risk as empirical modeling of apical responses.
It is less well understood that these exact same difficulties occur with BBDR models.
To link exposure with risk, a BBDR model must incorporate at least one biologic variable that is dose related.
BBDR modeling of low-dose effects is greatly complicated if the toxicant affects multiple intermediate steps in the disease process.
These problems are well illustrated by the experiences with BBDR models to date.
Many of the BBDR modeling efforts to date do not incorporate any biological data on one or more key intermediate variables.