concerns do arise, there are several avenues discussed below that can be followed
to isolate the factor(s) limiting the oral delivery of the compound.
In vitro experiments often provide valuable insight into the clearance
mechanisms for NCEs. The experiments that are most often employed in
tandem to understand bioavailablilty are determinations of compound
solubility, membrane permeability, and stability in subcellular fractions. The
subcellular fractions most often employed are plasma (for ester containing
compounds) and liver subcellular fractions with the addition of either NADPH
or UDPGA as cofactor. Hepatocytes are also a very usefulin vitrotool that
provides a more complete system for studying metabolism.
All of the assays mentioned can be set up in an automated, medium
throughput system, however, all require a specific assay to measure compound
concentration at the end of the assay that places significant limitations on
throughput. There have been recent attempts to solve the throughput problems
inherent with this type of assay by developing generic endpoint assays for
metabolic stability, but there has been no clear solution to date.
8.2.2 Prediction of Human Clearance
The rapid determination of pharmacokinetic parameters, solubility, perme-
ability, andin vitrostability in plasma or liver tissue can often provide a
reasonable explanation of the mechanisms limiting oral bioavailability. An
approach that is often used is to extrapolate thein vitrorate of metabolism to
estimate the hepatic clearance usingin vitro–in vivocorrelation methodology
(Houston and Carlile, 1997; Ito et al., 1998; Lave et al., 1997, 1999, 2002; Lin,
1998; Miners et al., 2006; Obach, 1999, 2001; Obach et al., 1997; Shiran et al.,
2006). These methods usein vitrokinetic parameters, usuallyVmax/Kmorin vitro
t1/2, to determine an intrinsic clearance, which is then scaled to hepatic
clearance using amount of tissue in thein vitroincubation, the weight of the
liver and the well-stirred model for hepatic clearance.
Care must be used when determiningin vitroparameters (Grime and Riley,
2006; Margolis and Obach, 2003; Miners et al., 2006; Obach, 1999; Obach
et al., 1997; Tran et al., 2002; Ziegler, 2002). The methods used to extrapolate
to thein vivosituation all assume that the drug concentration at the active site
of the CYP enzymes will be much less than theKmvalue. This is a reasonable
assumption for most drugs used clinically under typicalin vitroexperimental
conditions; however, experimental design, especiallyin vitro t1/2determinations,
should take this into account. The incorporation of protein binding corrections
in these calculations has been somewhat controversial. Clearance of ‘‘free
drug’’ was included in the first theoretical models for predicting hepatic
extraction fromin vitrodata (Rane et al., 1997) and was the favored method in
earlyin vitro–in vivocorrelation attempts. However, the inclusion of protein
binding into the scaling equations tends to underpredict actualin vivovalues
and generally better results are achieved when protein binding is left out of the
equation (Obach, 1997, 1999). The reasons for this could be (1) the intrahepatic
METABOLIC CLEARANCE 241