Principles and Practice of Pharmaceutical Medicine

trend for more efficient use of resources in early
clinical development.

8.1 The in vitro/in vivo prediction

The challenge is to predict systemic clearance,
volume of distribution and oral bioavailability in
humans from a combination ofin vitroandin vivo
preclinical data. If this prediction can become reli-
able, then phase I studies become more confirma-
tory. The use of human hepatocytes and isolated
enzymes can form a critical part of thein vitro
database.
Clearance of almost all drugs is by renal, meta-
bolic and/or biliary mechanisms. There are rare
exceptions, such as anesthetic gases that are
exhaled unchanged. However, in this chapter we
shall concentrate on the typical situation.
Physicochemical properties, especially lipophi-
licity, frequently govern the clearance route; lipo-
philicity is commonly measured as log D 7 : 4 , where
this variable equals log 10 ([drug in octanol]/[drug
in aqueous buffer]) at pH¼ 7 :4, in a closed system
at equilibrium. Generally, compounds with a log
D 7 : 4 value below 0 have significant renal clearance
values, whereas compounds with log D 7 : 4 values
above 0 will usually be eliminated principally by
hepatic metabolism (Smithet al., 1996). Molecular

size also has some effect on these clearance routes.

For example, compounds with molecular weights

greater than 400 Da are often eliminated through

the bile unchanged, whilst smaller lipophilic com-

pounds will generally be metabolized.

Elementary aspects of clearance

The common, clinical measurement of drug clear-

ance involves taking serial venous blood samples.

As time passes afterTmax(the time when drug

concentration reaches its peak), parent drug

concentrations continuously decline. Modeling of

drug disappearance is essentially a descriptive pro-

cess and requires actual human exposures. Unsa-

turated elimination mechanisms, in the absence of

drug sequestration, can be modeled as simple, first-

order elimination, using a constant (k) with units of

h
^1 ; plasma concentration (C) is then modeled by

equations of the general form:

C¼Ae
kt

whereAis the concentration of drug at time (t)0

(assuming that there was instantaneous and homo-

genousequilibration ofthe dose into thecirculating

compartment). As the number of compartments

increases, then so does the number of terms of

the form shown on the right-hand side of the equa-

tion shown above.

Theelimination ratealways has units of (mass/

time) for any elimination process. For first-order

processes, the elimination rate at any one moment

is represented by a tangent to the elimination curve

for any specified timetor drug concentrationC.

In contrast, zero-order elimination processes are

occasionally encountered. These usually represent

saturation by the drug of the elimination mechan-

ism(s). These ‘drug disappearance’ curves are

straight and thus described simply by:

C¼A
bt

where the elimination rate (b) does not change

with time or drug concentration. If followed for

long enough, most drugs that are subject to zero-

order elimination eventually fall to such low

Physicochemical
properties

(i) Microsomes

(ii) Hepatocytes

In vitro

rat

In vivo

rat

In vitro

human

In vivo

human

Second

mammalian

species Effect

(pharmacokinetic–pharmacodynamic

models)

Bioanalysis Second

mammalian

Species

Figure 8.1 General scheme showing the pharmacoki-
netic prediction pathway from physicochemical proper-
ties to human drug response viain vitroandin vivo
studies in laboratory animals

80 CH8 PHASE I: THE FIRST OPPORTUNITY FOR EXTRAPOLATION FROM ANIMAL DATA