Using Emerging Science and Technologies to Address Persistent and Future Environmental Challenges - Science For Environmental Protection: The Road Ahead - page 52

cate the potential for future toxicity, but many intervening biologic processes
may abrogate downstream responses, so the fact that a particular pathway is
activated by a chemical does not necessarily mean that the same will occur in
vivo. It will be important for high-throughput screening approaches to consider
multiple time points for analysis.
 Dose-response assessment determined in vitro may be difficult to cor-
relate with in vivo responses and administered doses. Relating dose rate (in mil-
ligrams per kilogram per day) in vivo at specific tissues to cell-culture concen-
trations tested in vitro is extremely difficult and requires detailed knowledge of
the absorption, distribution, metabolism, and excretion of a xenobiotic after in
vivo exposure. It also requires knowledge about protein binding to plasma and
intracellular proteins, lipid portioning, tissue-specific activation, and detoxifica-
tion for interpretation of the relevance of an in vitro cell concentration to a tar-
get-tissue concentration after in vivo administration. Thus, physiologically based
pharmacokinetic modeling, which will require some in vivo data, will continue
to be an important part of hazard evaluation and risk assessment for chemicals
that are identified as being potentially of concern on the basis of in vitro screen-
ing assays. Although advances in in vitro toxicity assessment continue to im-
prove and will certainly decrease the number of animals required for in vivo
testing, it is unlikely that in vitro tests will fully replace the need for in vivo
animal testing for understanding the pharmacokinetics and pharmacodynamics
of toxic substances because of the complex interplay between tissues and organs
that are ultimately critical determinates of a toxic response.
The importance of those concepts was recently illustrated in some model-
ing studies of EPA ToxCast data. In the first phase of the ToxCast program,
EPA scientists used hundreds of in vitro assays to screen a library of agricultural
and industrial chemicals to identify cellular pathways and processes that were
modified by specific chemicals; they intended to use the data to set priorities
among chemicals for further testing (Judson et al. 2010). However, the potency
of a chemical in an in vitro assay may or may not reflect its biologic potency in
vivo because of differences in bioavailability, clearance, and exposure (Blaau-
boer 2010). Scientists at the Hamner Institute, in collaboration with EPA scien-
tists, recently developed pharmacokinetic and pharmacodynamic models that
incorporate human dosimetry and exposure data with the ToxCast high-
throughput in vitro screening data (Rotroff et al. 2010; Wetmore et al. 2012).
Their results demonstrated that incorporation of dosimetry and exposure infor-
mation is critically important for improving priority-setting for further testing
and for evaluating the potential human health effects at relevant exposures.
EPA scientists have played a leading role in the new approaches, and it
will be important for them to continue to lead the way in both computational and
systems toxicology in the future. With further improvements, such as inclusion
of human dosimetric and exposure data, high-throughput in vitro assays for
screening of new chemical entities for potentially hazardous properties will