which may not affect the results of thein vitroscreening assays, but may have a
major effect on the effective concentration of the inducer at its site of induction
in vivo. For other CYP enzymes, the correlation between induction potential
predicted byin vitroassays, and actual induction seenin vivois thus far less
firmly established, primarily because of a more limited number of studies and a
smaller available database.
17.3 Final Comments
Clinically relevant induction of CYP enzymes occurs during multiple dosing of
the inducing drug and is a dose- and time-dependent phenomenon. We have
described several short-term methods in this chapter, bothin vivoandin vitro,
for determining the potential of a candidate molecule to induce CYP3A4. A
number of criteria for the usefulness of data from such methods, especiallyin
vitroassays (Bjornsson et al., 2003; Tucker et al., 2001), have been proposed.
The use of both positive and negative controls are crucial in these assays.
Compounds should be tested at concentrations covering the targeted human
Cmax, and dose–response curves should be generated when possible.
Reproducibility of cultured human hepatocyte-based assays is an important
issue, and ideally at least three different replicates from three different human
liver sources should be used for each treatment group. We have also suggested
that a response of at least 40% of that produced by the positive control
compounds level be considered as the criterion for a positive induction
response in thein vitroassays.
In vitroand animal models possess some advantages over the more direct,
but also more expensive clinical studies. However, although they can provide a
qualitative or semiquantitative assessment of induction, these preclinical
methods cannot as yet predict the degree of CYP induction in humans by a
candidate molecule with quantitative certainty. A recent study by Kato et al.
(2005) compared literature data onin vitroandin vivoenzyme induction with
their data obtained with human hepatocytesin vitro, normalized for non-
protein bound mean concentrations of inducers. They generally showed that
maximum induction ratios produced by inducers inin vitrostudies were higher
than those seen inin vivostudies. The authors provide evidence that it is
possible to apply pharmacokinetic corrections to accurately predict, on a
quantitative basis, human CYP3A inductionin vivousing data obtained with
human hepatocytesin vitro. Using immortalized human hepatocyte Fa2N-4
cells and efficacious free human plasma concentrations, Ripp et al. recently
have reported good correlations between percentage AUC changes and relative
induction scores with midazolam or ethinyl estradiol as CYP3A4 substrates
(Ripp et al., 2006).
Finally, in addition to mediating the process of CYP3A4 induction in
humans, the PXR receptor is also responsible for the regulation of expression
of Phase II conjugating enzymes such as UGT1A1, and it also upregulates
FINAL COMMENTS 565